diff --git a/.gitattributes b/.gitattributes
new file mode 100644
index 0000000000000000000000000000000000000000..a09db2ea4d1bd1fc1c09f6fca45e8e4054953535
--- /dev/null
+++ b/.gitattributes
@@ -0,0 +1,37 @@
+*.7z filter=lfs diff=lfs merge=lfs -text
+*.arrow filter=lfs diff=lfs merge=lfs -text
+*.bin filter=lfs diff=lfs merge=lfs -text
+*.bz2 filter=lfs diff=lfs merge=lfs -text
+*.ckpt filter=lfs diff=lfs merge=lfs -text
+*.ftz filter=lfs diff=lfs merge=lfs -text
+*.gz filter=lfs diff=lfs merge=lfs -text
+*.h5 filter=lfs diff=lfs merge=lfs -text
+*.joblib filter=lfs diff=lfs merge=lfs -text
+*.lfs.* filter=lfs diff=lfs merge=lfs -text
+*.mlmodel filter=lfs diff=lfs merge=lfs -text
+*.model filter=lfs diff=lfs merge=lfs -text
+*.msgpack filter=lfs diff=lfs merge=lfs -text
+*.npy filter=lfs diff=lfs merge=lfs -text
+*.npz filter=lfs diff=lfs merge=lfs -text
+*.onnx filter=lfs diff=lfs merge=lfs -text
+*.ot filter=lfs diff=lfs merge=lfs -text
+*.parquet filter=lfs diff=lfs merge=lfs -text
+*.pb filter=lfs diff=lfs merge=lfs -text
+*.pickle filter=lfs diff=lfs merge=lfs -text
+*.pkl filter=lfs diff=lfs merge=lfs -text
+*.pt filter=lfs diff=lfs merge=lfs -text
+*.pth filter=lfs diff=lfs merge=lfs -text
+*.rar filter=lfs diff=lfs merge=lfs -text
+*.safetensors filter=lfs diff=lfs merge=lfs -text
+saved_model/**/* filter=lfs diff=lfs merge=lfs -text
+*.tar.* filter=lfs diff=lfs merge=lfs -text
+*.tar filter=lfs diff=lfs merge=lfs -text
+*.tflite filter=lfs diff=lfs merge=lfs -text
+*.tgz filter=lfs diff=lfs merge=lfs -text
+*.wasm filter=lfs diff=lfs merge=lfs -text
+*.xz filter=lfs diff=lfs merge=lfs -text
+*.zip filter=lfs diff=lfs merge=lfs -text
+*.zst filter=lfs diff=lfs merge=lfs -text
+*tfevents* filter=lfs diff=lfs merge=lfs -text
+tokenizer.json filter=lfs diff=lfs merge=lfs -text
+model.safetensors.index.json filter=lfs diff=lfs merge=lfs -text
diff --git a/.quant_summary.txt b/.quant_summary.txt
new file mode 100644
index 0000000000000000000000000000000000000000..81d52f95fa432e361fd08d64a021b7e1cd629b18
--- /dev/null
+++ b/.quant_summary.txt
@@ -0,0 +1,40972 @@
+model.embed_tokens.weight_quantizer                                              TensorQuantizer(disabled)
+model.embed_tokens.input_quantizer                                               HardDisabledTensorQuantizer(disabled)
+model.embed_tokens.output_quantizer                                              TensorQuantizer(disabled)
+model.layers.0.self_attn.q_a_proj.input_quantizer                                TensorQuantizer(disabled)
+model.layers.0.self_attn.q_a_proj.output_quantizer                               TensorQuantizer(disabled)
+model.layers.0.self_attn.q_a_proj.weight_quantizer                               TensorQuantizer(disabled)
+model.layers.0.self_attn.q_b_proj.input_quantizer                                TensorQuantizer(disabled)
+model.layers.0.self_attn.q_b_proj.output_quantizer                               TensorQuantizer(disabled)
+model.layers.0.self_attn.q_b_proj.weight_quantizer                               TensorQuantizer(disabled)
+model.layers.0.self_attn.kv_a_proj_with_mqa.input_quantizer                      TensorQuantizer(disabled)
+model.layers.0.self_attn.kv_a_proj_with_mqa.output_quantizer                     TensorQuantizer(disabled)
+model.layers.0.self_attn.kv_a_proj_with_mqa.weight_quantizer                     TensorQuantizer(disabled)
+model.layers.0.self_attn.kv_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.0.self_attn.kv_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.0.self_attn.kv_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.0.self_attn.o_proj.input_quantizer                                  TensorQuantizer(disabled)
+model.layers.0.self_attn.o_proj.output_quantizer                                 TensorQuantizer(disabled)
+model.layers.0.self_attn.o_proj.weight_quantizer                                 TensorQuantizer(disabled)
+model.layers.0.self_attn.indexer.wq_b.input_quantizer                            TensorQuantizer(disabled)
+model.layers.0.self_attn.indexer.wq_b.output_quantizer                           TensorQuantizer(disabled)
+model.layers.0.self_attn.indexer.wq_b.weight_quantizer                           TensorQuantizer(disabled)
+model.layers.0.self_attn.indexer.wk.input_quantizer                              TensorQuantizer(disabled)
+model.layers.0.self_attn.indexer.wk.output_quantizer                             TensorQuantizer(disabled)
+model.layers.0.self_attn.indexer.wk.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.0.self_attn.indexer.k_norm.input_quantizer                          TensorQuantizer(disabled)
+model.layers.0.self_attn.indexer.k_norm.output_quantizer                         TensorQuantizer(disabled)
+model.layers.0.self_attn.indexer.weights_proj.input_quantizer                    TensorQuantizer(disabled)
+model.layers.0.self_attn.indexer.weights_proj.output_quantizer                   TensorQuantizer(disabled)
+model.layers.0.self_attn.indexer.weights_proj.weight_quantizer                   TensorQuantizer(disabled)
+model.layers.0.self_attn.q_bmm_quantizer                                         TensorQuantizer(disabled)
+model.layers.0.self_attn.k_bmm_quantizer                                         TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.0.self_attn.v_bmm_quantizer                                         TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.0.self_attn.softmax_quantizer                                       TensorQuantizer(disabled)
+model.layers.0.mlp.gate_proj.input_quantizer                                     TensorQuantizer(disabled)
+model.layers.0.mlp.gate_proj.output_quantizer                                    TensorQuantizer(disabled)
+model.layers.0.mlp.gate_proj.weight_quantizer                                    TensorQuantizer(disabled)
+model.layers.0.mlp.up_proj.input_quantizer                                       TensorQuantizer(disabled)
+model.layers.0.mlp.up_proj.output_quantizer                                      TensorQuantizer(disabled)
+model.layers.0.mlp.up_proj.weight_quantizer                                      TensorQuantizer(disabled)
+model.layers.0.mlp.down_proj.input_quantizer                                     TensorQuantizer(disabled)
+model.layers.0.mlp.down_proj.output_quantizer                                    TensorQuantizer(disabled)
+model.layers.0.mlp.down_proj.weight_quantizer                                    TensorQuantizer(disabled)
+model.layers.1.self_attn.q_a_proj.input_quantizer                                TensorQuantizer(disabled)
+model.layers.1.self_attn.q_a_proj.output_quantizer                               TensorQuantizer(disabled)
+model.layers.1.self_attn.q_a_proj.weight_quantizer                               TensorQuantizer(disabled)
+model.layers.1.self_attn.q_b_proj.input_quantizer                                TensorQuantizer(disabled)
+model.layers.1.self_attn.q_b_proj.output_quantizer                               TensorQuantizer(disabled)
+model.layers.1.self_attn.q_b_proj.weight_quantizer                               TensorQuantizer(disabled)
+model.layers.1.self_attn.kv_a_proj_with_mqa.input_quantizer                      TensorQuantizer(disabled)
+model.layers.1.self_attn.kv_a_proj_with_mqa.output_quantizer                     TensorQuantizer(disabled)
+model.layers.1.self_attn.kv_a_proj_with_mqa.weight_quantizer                     TensorQuantizer(disabled)
+model.layers.1.self_attn.kv_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.1.self_attn.kv_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.1.self_attn.kv_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.1.self_attn.o_proj.input_quantizer                                  TensorQuantizer(disabled)
+model.layers.1.self_attn.o_proj.output_quantizer                                 TensorQuantizer(disabled)
+model.layers.1.self_attn.o_proj.weight_quantizer                                 TensorQuantizer(disabled)
+model.layers.1.self_attn.indexer.wq_b.input_quantizer                            TensorQuantizer(disabled)
+model.layers.1.self_attn.indexer.wq_b.output_quantizer                           TensorQuantizer(disabled)
+model.layers.1.self_attn.indexer.wq_b.weight_quantizer                           TensorQuantizer(disabled)
+model.layers.1.self_attn.indexer.wk.input_quantizer                              TensorQuantizer(disabled)
+model.layers.1.self_attn.indexer.wk.output_quantizer                             TensorQuantizer(disabled)
+model.layers.1.self_attn.indexer.wk.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.1.self_attn.indexer.k_norm.input_quantizer                          TensorQuantizer(disabled)
+model.layers.1.self_attn.indexer.k_norm.output_quantizer                         TensorQuantizer(disabled)
+model.layers.1.self_attn.indexer.weights_proj.input_quantizer                    TensorQuantizer(disabled)
+model.layers.1.self_attn.indexer.weights_proj.output_quantizer                   TensorQuantizer(disabled)
+model.layers.1.self_attn.indexer.weights_proj.weight_quantizer                   TensorQuantizer(disabled)
+model.layers.1.self_attn.q_bmm_quantizer                                         TensorQuantizer(disabled)
+model.layers.1.self_attn.k_bmm_quantizer                                         TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.1.self_attn.v_bmm_quantizer                                         TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.1.self_attn.softmax_quantizer                                       TensorQuantizer(disabled)
+model.layers.1.mlp.gate_proj.input_quantizer                                     TensorQuantizer(disabled)
+model.layers.1.mlp.gate_proj.output_quantizer                                    TensorQuantizer(disabled)
+model.layers.1.mlp.gate_proj.weight_quantizer                                    TensorQuantizer(disabled)
+model.layers.1.mlp.up_proj.input_quantizer                                       TensorQuantizer(disabled)
+model.layers.1.mlp.up_proj.output_quantizer                                      TensorQuantizer(disabled)
+model.layers.1.mlp.up_proj.weight_quantizer                                      TensorQuantizer(disabled)
+model.layers.1.mlp.down_proj.input_quantizer                                     TensorQuantizer(disabled)
+model.layers.1.mlp.down_proj.output_quantizer                                    TensorQuantizer(disabled)
+model.layers.1.mlp.down_proj.weight_quantizer                                    TensorQuantizer(disabled)
+model.layers.2.self_attn.q_a_proj.input_quantizer                                TensorQuantizer(disabled)
+model.layers.2.self_attn.q_a_proj.output_quantizer                               TensorQuantizer(disabled)
+model.layers.2.self_attn.q_a_proj.weight_quantizer                               TensorQuantizer(disabled)
+model.layers.2.self_attn.q_b_proj.input_quantizer                                TensorQuantizer(disabled)
+model.layers.2.self_attn.q_b_proj.output_quantizer                               TensorQuantizer(disabled)
+model.layers.2.self_attn.q_b_proj.weight_quantizer                               TensorQuantizer(disabled)
+model.layers.2.self_attn.kv_a_proj_with_mqa.input_quantizer                      TensorQuantizer(disabled)
+model.layers.2.self_attn.kv_a_proj_with_mqa.output_quantizer                     TensorQuantizer(disabled)
+model.layers.2.self_attn.kv_a_proj_with_mqa.weight_quantizer                     TensorQuantizer(disabled)
+model.layers.2.self_attn.kv_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.2.self_attn.kv_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.2.self_attn.kv_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.2.self_attn.o_proj.input_quantizer                                  TensorQuantizer(disabled)
+model.layers.2.self_attn.o_proj.output_quantizer                                 TensorQuantizer(disabled)
+model.layers.2.self_attn.o_proj.weight_quantizer                                 TensorQuantizer(disabled)
+model.layers.2.self_attn.indexer.wq_b.input_quantizer                            TensorQuantizer(disabled)
+model.layers.2.self_attn.indexer.wq_b.output_quantizer                           TensorQuantizer(disabled)
+model.layers.2.self_attn.indexer.wq_b.weight_quantizer                           TensorQuantizer(disabled)
+model.layers.2.self_attn.indexer.wk.input_quantizer                              TensorQuantizer(disabled)
+model.layers.2.self_attn.indexer.wk.output_quantizer                             TensorQuantizer(disabled)
+model.layers.2.self_attn.indexer.wk.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.2.self_attn.indexer.k_norm.input_quantizer                          TensorQuantizer(disabled)
+model.layers.2.self_attn.indexer.k_norm.output_quantizer                         TensorQuantizer(disabled)
+model.layers.2.self_attn.indexer.weights_proj.input_quantizer                    TensorQuantizer(disabled)
+model.layers.2.self_attn.indexer.weights_proj.output_quantizer                   TensorQuantizer(disabled)
+model.layers.2.self_attn.indexer.weights_proj.weight_quantizer                   TensorQuantizer(disabled)
+model.layers.2.self_attn.q_bmm_quantizer                                         TensorQuantizer(disabled)
+model.layers.2.self_attn.k_bmm_quantizer                                         TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.2.self_attn.v_bmm_quantizer                                         TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.2.self_attn.softmax_quantizer                                       TensorQuantizer(disabled)
+model.layers.2.mlp.gate_proj.input_quantizer                                     TensorQuantizer(disabled)
+model.layers.2.mlp.gate_proj.output_quantizer                                    TensorQuantizer(disabled)
+model.layers.2.mlp.gate_proj.weight_quantizer                                    TensorQuantizer(disabled)
+model.layers.2.mlp.up_proj.input_quantizer                                       TensorQuantizer(disabled)
+model.layers.2.mlp.up_proj.output_quantizer                                      TensorQuantizer(disabled)
+model.layers.2.mlp.up_proj.weight_quantizer                                      TensorQuantizer(disabled)
+model.layers.2.mlp.down_proj.input_quantizer                                     TensorQuantizer(disabled)
+model.layers.2.mlp.down_proj.output_quantizer                                    TensorQuantizer(disabled)
+model.layers.2.mlp.down_proj.weight_quantizer                                    TensorQuantizer(disabled)
+model.layers.3.self_attn.q_a_proj.input_quantizer                                TensorQuantizer(disabled)
+model.layers.3.self_attn.q_a_proj.output_quantizer                               TensorQuantizer(disabled)
+model.layers.3.self_attn.q_a_proj.weight_quantizer                               TensorQuantizer(disabled)
+model.layers.3.self_attn.q_b_proj.input_quantizer                                TensorQuantizer(disabled)
+model.layers.3.self_attn.q_b_proj.output_quantizer                               TensorQuantizer(disabled)
+model.layers.3.self_attn.q_b_proj.weight_quantizer                               TensorQuantizer(disabled)
+model.layers.3.self_attn.kv_a_proj_with_mqa.input_quantizer                      TensorQuantizer(disabled)
+model.layers.3.self_attn.kv_a_proj_with_mqa.output_quantizer                     TensorQuantizer(disabled)
+model.layers.3.self_attn.kv_a_proj_with_mqa.weight_quantizer                     TensorQuantizer(disabled)
+model.layers.3.self_attn.kv_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.3.self_attn.kv_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.3.self_attn.kv_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.3.self_attn.o_proj.input_quantizer                                  TensorQuantizer(disabled)
+model.layers.3.self_attn.o_proj.output_quantizer                                 TensorQuantizer(disabled)
+model.layers.3.self_attn.o_proj.weight_quantizer                                 TensorQuantizer(disabled)
+model.layers.3.self_attn.q_bmm_quantizer                                         TensorQuantizer(disabled)
+model.layers.3.self_attn.k_bmm_quantizer                                         TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.3.self_attn.v_bmm_quantizer                                         TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.3.self_attn.softmax_quantizer                                       TensorQuantizer(disabled)
+model.layers.3.mlp.experts.gate_up_proj_input_quantizer                          TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.10e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.0                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.12e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.1                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.46e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.2                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.93e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.3                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.82e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.4                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.5                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.87e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.6                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.7                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.54e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.8                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.97e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.9                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.10                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.76e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.11                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.89e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.12                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.92e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.13                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.14                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.15                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.92e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.16                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.17                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.18                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.85e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.19                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.91e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.20                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.21                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.22                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.75e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.23                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.24                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.25                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.26                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.27                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.28                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.29                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.77e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.30                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.31                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.32                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.33                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.75e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.34                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.87e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.35                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.36                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.37                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.38                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.00e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.39                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.21e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.40                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.38e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.41                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.02e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.42                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.05e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.43                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.03e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.44                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.45                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.46                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.47                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.48                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.49                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.50                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.51                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.35e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.52                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.53                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.54                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.55                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.56                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.57                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.58                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.79e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.59                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.60                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.81e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.61                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.62                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.86e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.63                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.64                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.65                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.66                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.22e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.67                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.07e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.68                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.83e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.69                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.70                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.71                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.22e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.72                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.73                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.74                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.85e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.75                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.76                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.78e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.77                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.69e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.78                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.80e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.79                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.54e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.80                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.81                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.82                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.31e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.83                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.90e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.84                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.85                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.23e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.86                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.87                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.01e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.88                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.29e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.89                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.99e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.90                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.91                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.86e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.92                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.97e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.93                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.78e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.94                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.06e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.95                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.52e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.96                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.89e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.97                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.56e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.98                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.18e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.99                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.100                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.101                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.34e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.102                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.35e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.103                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.104                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.08e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.105                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.106                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.107                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.108                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.109                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.78e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.110                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.05e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.111                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.112                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.113                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.29e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.114                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.95e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.115                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.116                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.117                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.118                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.119                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.120                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.121                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.122                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.123                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.83e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.124                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.79e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.125                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.99e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.126                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.84e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.127                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.128                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.129                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.130                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.03e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.131                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.79e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.132                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.76e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.133                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.01e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.134                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.77e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.135                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.17e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.136                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.137                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.138                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.139                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.80e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.140                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.141                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.44e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.142                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.95e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.143                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.22e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.144                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.79e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.145                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.146                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.147                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.77e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.148                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.01e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.149                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.150                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.80e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.151                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.152                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.13e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.153                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.154                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.155                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.09e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.156                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.09e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.157                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.158                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.159                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.47e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.160                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.161                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.64e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.162                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.163                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.164                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.165                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.166                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.167                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.168                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.97e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.169                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.170                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.171                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.95e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.172                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.173                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.90e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.174                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.175                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.176                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.97e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.177                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.77e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.178                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.83e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.179                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.180                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.181                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.182                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.30e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.183                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.184                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.04e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.185                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.96e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.186                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.80e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.187                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.99e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.188                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.189                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.95e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.190                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.191                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.192                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.193                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.83e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.194                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.195                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.196                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.83e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.197                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.198                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.199                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.200                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.95e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.201                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.202                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.00e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.203                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.204                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.95e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.205                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.206                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.78e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.207                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.28e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.208                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.99e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.209                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.210                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.99e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.211                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.80e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.212                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.213                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.98e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.214                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.215                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.77e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.216                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.28e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.217                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.218                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.219                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.22e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.220                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.221                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.19e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.222                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.24e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.223                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.224                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.225                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.226                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.54e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.227                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.228                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.95e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.229                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.12e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.230                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.20e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.231                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.54e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.232                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.233                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.234                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.235                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.236                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.95e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.237                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.238                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.239                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.240                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.241                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.242                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.243                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.244                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.76e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.245                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.46e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.246                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.54e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.247                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.39e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.248                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.79e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.249                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.77e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.250                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e+00 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.251                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.252                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.253                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.254                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.255                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.13e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_input_quantizer                             TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e+00 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.0                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.1                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.2                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.37e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.3                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.4                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.5                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.6                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.7                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.8                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.77e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.9                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.10                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.11                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.12                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.13                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.14                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.15                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.16                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.17                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.18                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.05e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.19                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.20                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.21                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.22                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.23                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.24                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.25                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.26                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.27                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.28                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.25e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.29                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.30                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.31                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.32                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.33                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.34                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.35                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.36                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.37                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.38                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.39                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.40                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.41                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.30e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.42                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.43                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.77e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.44                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.45                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.46                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.01e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.47                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.48                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.49                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.89e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.50                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.51                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.52                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.53                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.30e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.54                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.55                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.56                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.57                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.58                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.59                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.60                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.34e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.61                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.62                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.11e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.63                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.64                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.65                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.66                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.67                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.68                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.69                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.70                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.71                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.72                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.73                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.74                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.75                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.76                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.77                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.52e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.78                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.79                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.80                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.81                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.82                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.83                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.84                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.85                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.86                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.15e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.87                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.87e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.88                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.89                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.71e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.90                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.91                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.92                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.58e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.93                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.94                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.95                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.96                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.97                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.98                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.99                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.100                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.101                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.36e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.102                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.103                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.104                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.105                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.106                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.107                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.108                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.109                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.110                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.111                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.112                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.113                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.114                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.115                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.116                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.117                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.40e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.118                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.119                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.120                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.121                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.122                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.25e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.123                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.124                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.125                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.126                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.127                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.128                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.129                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.130                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.131                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.132                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.133                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.134                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.135                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.136                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.137                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.80e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.138                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.139                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.140                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.141                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.142                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.86e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.143                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.09e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.144                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.145                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.146                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.147                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.76e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.148                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.85e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.149                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.150                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.151                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.152                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.153                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.154                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.155                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.156                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.157                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.158                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.159                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.160                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.161                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.162                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.163                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.164                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.165                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.166                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.35e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.167                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.168                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.169                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.25e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.170                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.171                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.172                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.173                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.174                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.175                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.176                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.177                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.178                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.179                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.180                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.181                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.182                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.183                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.184                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.25e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.185                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.186                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.187                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.188                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.189                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.190                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.191                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.192                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.193                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.194                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.44e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.195                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.196                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.197                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.91e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.198                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.199                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.200                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.201                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.202                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.203                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.204                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.205                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.206                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.207                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.208                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.209                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.210                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.211                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.212                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.213                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.214                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.40e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.215                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.216                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.217                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.218                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.30e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.219                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.83e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.220                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.221                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.222                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.223                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.224                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.225                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.226                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.227                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.228                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.229                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.08e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.230                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.95e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.231                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.232                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.233                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.234                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.235                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.236                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.83e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.237                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.238                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.239                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.15e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.240                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.241                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.242                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.243                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.244                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.245                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.59e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.246                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.02e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.247                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.248                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.249                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.250                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.12e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.251                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.252                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.253                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.254                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.experts.down_proj_weight_quantizers.255                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.3.mlp.shared_experts.gate_proj.input_quantizer                      TensorQuantizer(disabled)
+model.layers.3.mlp.shared_experts.gate_proj.output_quantizer                     TensorQuantizer(disabled)
+model.layers.3.mlp.shared_experts.gate_proj.weight_quantizer                     TensorQuantizer(disabled)
+model.layers.3.mlp.shared_experts.up_proj.input_quantizer                        TensorQuantizer(disabled)
+model.layers.3.mlp.shared_experts.up_proj.output_quantizer                       TensorQuantizer(disabled)
+model.layers.3.mlp.shared_experts.up_proj.weight_quantizer                       TensorQuantizer(disabled)
+model.layers.3.mlp.shared_experts.down_proj.input_quantizer                      TensorQuantizer(disabled)
+model.layers.3.mlp.shared_experts.down_proj.output_quantizer                     TensorQuantizer(disabled)
+model.layers.3.mlp.shared_experts.down_proj.weight_quantizer                     TensorQuantizer(disabled)
+model.layers.4.self_attn.q_a_proj.input_quantizer                                TensorQuantizer(disabled)
+model.layers.4.self_attn.q_a_proj.output_quantizer                               TensorQuantizer(disabled)
+model.layers.4.self_attn.q_a_proj.weight_quantizer                               TensorQuantizer(disabled)
+model.layers.4.self_attn.q_b_proj.input_quantizer                                TensorQuantizer(disabled)
+model.layers.4.self_attn.q_b_proj.output_quantizer                               TensorQuantizer(disabled)
+model.layers.4.self_attn.q_b_proj.weight_quantizer                               TensorQuantizer(disabled)
+model.layers.4.self_attn.kv_a_proj_with_mqa.input_quantizer                      TensorQuantizer(disabled)
+model.layers.4.self_attn.kv_a_proj_with_mqa.output_quantizer                     TensorQuantizer(disabled)
+model.layers.4.self_attn.kv_a_proj_with_mqa.weight_quantizer                     TensorQuantizer(disabled)
+model.layers.4.self_attn.kv_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.4.self_attn.kv_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.4.self_attn.kv_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.4.self_attn.o_proj.input_quantizer                                  TensorQuantizer(disabled)
+model.layers.4.self_attn.o_proj.output_quantizer                                 TensorQuantizer(disabled)
+model.layers.4.self_attn.o_proj.weight_quantizer                                 TensorQuantizer(disabled)
+model.layers.4.self_attn.q_bmm_quantizer                                         TensorQuantizer(disabled)
+model.layers.4.self_attn.k_bmm_quantizer                                         TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.4.self_attn.v_bmm_quantizer                                         TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.4.self_attn.softmax_quantizer                                       TensorQuantizer(disabled)
+model.layers.4.mlp.experts.gate_up_proj_input_quantizer                          TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.64e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.0                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.1                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.2                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.3                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.4                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.5                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.6                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.7                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.8                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.00e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.9                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.10                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.11                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.75e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.12                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.13                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.14                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.15                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.16                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.17                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.00e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.18                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.19                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.20                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.21                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.22                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.23                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.10e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.24                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.25                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.26                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.15e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.27                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.28                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.98e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.29                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.90e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.30                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.31                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.32                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.33                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.34                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.35                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.36                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.76e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.37                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.38                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.73e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.39                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.97e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.40                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.41                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.42                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.43                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.95e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.44                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.45                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.46                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.95e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.47                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.48                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.49                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.50                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.44e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.51                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.52                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.53                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.54                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.93e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.55                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.56                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.57                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.58                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.59                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.60                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.61                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.62                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.63                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.64                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.65                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.66                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.67                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.68                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.69                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.70                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.06e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.71                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.25e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.72                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.73                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.74                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.03e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.75                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.76                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.77                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.62e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.78                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.79                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.90e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.80                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.81                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.82                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.83                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.84                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.85                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.86                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.87                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.88                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.93e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.89                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.95e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.90                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.91                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.92                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.93                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.94                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.95                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.96                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.97                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.98                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.71e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.99                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.100                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.101                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.102                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.103                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.104                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.105                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.106                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.107                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.83e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.108                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.109                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.110                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.111                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.112                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.98e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.113                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.114                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.115                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.116                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.117                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.118                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.119                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.120                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.121                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.122                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.123                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.78e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.124                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.125                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.126                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.11e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.127                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.128                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.129                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.130                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.84e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.131                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.132                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.27e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.133                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.134                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.135                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.96e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.136                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.94e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.137                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.138                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.139                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.92e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.140                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.75e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.141                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.73e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.142                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.143                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.144                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.145                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.146                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.147                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.71e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.148                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.149                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.150                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.151                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.152                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.153                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.154                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.155                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.156                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.157                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.158                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.159                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.160                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.16e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.161                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.162                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.43e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.163                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.164                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.165                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.166                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.167                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.168                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.47e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.169                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.170                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.171                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.89e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.172                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.173                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.174                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.175                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.176                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.177                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.75e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.178                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.87e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.179                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.180                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.181                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.73e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.182                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.183                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.184                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.185                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.80e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.186                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.87e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.187                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.75e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.188                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.189                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.190                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.191                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.192                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.193                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.96e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.194                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.01e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.195                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.196                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.197                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.198                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.199                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.50e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.200                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.201                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.202                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.203                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.87e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.204                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.205                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.206                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.207                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.208                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.209                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.45e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.210                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.211                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.212                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.213                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.214                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.52e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.215                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.83e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.216                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.83e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.217                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.218                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.219                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.07e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.220                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.221                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.222                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.223                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.224                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.225                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.226                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.91e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.227                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.228                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.229                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.97e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.230                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.231                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.232                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.233                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.234                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.235                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.85e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.236                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.237                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.97e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.238                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.87e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.239                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.240                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.241                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.86e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.242                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.243                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.244                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.245                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.83e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.246                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.247                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.75e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.248                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.86e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.249                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.250                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.93e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.251                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.252                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.77e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.253                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.254                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.255                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_input_quantizer                             TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e+00 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.0                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.1                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.2                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.3                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.4                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.5                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.6                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.7                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.8                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.20e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.9                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.10                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.11                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.12                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.13                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.14                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.15                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.16                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.17                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.18                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.75e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.19                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.20                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.21                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.22                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.23                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.77e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.24                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.25                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.26                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.27                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.28                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.29                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.30                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.02e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.31                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.32                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.33                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.34                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.35                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.36                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.37                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.38                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.79e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.39                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.40                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.41                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.42                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.84e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.43                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.44                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.45                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.46                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.47                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.78e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.48                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.49                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.50                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.51                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.52                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.53                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.54                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.55                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.56                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.57                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.58                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.59                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.60                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.61                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.62                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.63                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.64                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.65                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.66                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.67                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.68                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.69                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.85e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.70                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.71                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.72                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.73                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.74                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.75                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.76                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.77                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.82e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.78                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.79                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.80                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.81                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.82                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.83                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.84                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.04e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.85                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.86                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.87                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.88                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.89                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.90                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.91                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.92                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.93                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.94                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.95                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.96                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.97                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.98                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.99                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.100                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.101                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.102                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.103                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.104                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.105                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.106                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.107                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.108                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.109                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.110                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.111                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.112                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.113                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.114                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.115                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.116                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.117                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.118                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.92e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.119                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.120                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.121                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.122                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.123                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.124                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.125                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.23e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.126                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.83e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.127                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.10e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.128                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.73e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.129                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.130                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.131                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.132                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.133                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.134                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.135                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.136                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.137                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.138                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.139                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.140                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.141                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.142                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.143                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.144                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.145                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.146                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.147                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.37e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.148                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.149                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.150                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.11e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.151                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.152                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.93e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.153                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.154                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.85e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.155                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.156                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.157                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.158                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.159                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.160                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.161                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.162                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.163                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.164                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.165                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.166                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.167                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.168                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.70e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.169                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.170                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.171                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.172                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.173                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.174                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.175                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.176                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.177                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.178                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.179                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.180                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.181                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.182                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.183                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.184                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.91e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.185                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.186                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.187                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.188                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.189                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.190                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.191                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.192                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.193                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.17e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.194                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.195                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.196                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.197                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.198                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.199                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.19e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.200                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.201                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.202                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.203                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.204                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.205                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.206                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.09e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.207                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.78e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.208                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.209                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.210                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.211                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.212                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.213                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.214                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.215                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.216                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.217                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.218                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.219                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.220                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.221                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.222                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.223                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.224                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.225                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.226                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.227                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.228                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.229                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.230                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.231                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.232                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.233                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.234                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.235                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.236                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.237                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.238                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.239                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.240                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.241                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.82e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.242                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.243                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.244                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.245                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.246                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.247                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.248                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.249                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.250                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.251                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.252                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.253                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.254                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.experts.down_proj_weight_quantizers.255                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.4.mlp.shared_experts.gate_proj.input_quantizer                      TensorQuantizer(disabled)
+model.layers.4.mlp.shared_experts.gate_proj.output_quantizer                     TensorQuantizer(disabled)
+model.layers.4.mlp.shared_experts.gate_proj.weight_quantizer                     TensorQuantizer(disabled)
+model.layers.4.mlp.shared_experts.up_proj.input_quantizer                        TensorQuantizer(disabled)
+model.layers.4.mlp.shared_experts.up_proj.output_quantizer                       TensorQuantizer(disabled)
+model.layers.4.mlp.shared_experts.up_proj.weight_quantizer                       TensorQuantizer(disabled)
+model.layers.4.mlp.shared_experts.down_proj.input_quantizer                      TensorQuantizer(disabled)
+model.layers.4.mlp.shared_experts.down_proj.output_quantizer                     TensorQuantizer(disabled)
+model.layers.4.mlp.shared_experts.down_proj.weight_quantizer                     TensorQuantizer(disabled)
+model.layers.5.self_attn.q_a_proj.input_quantizer                                TensorQuantizer(disabled)
+model.layers.5.self_attn.q_a_proj.output_quantizer                               TensorQuantizer(disabled)
+model.layers.5.self_attn.q_a_proj.weight_quantizer                               TensorQuantizer(disabled)
+model.layers.5.self_attn.q_b_proj.input_quantizer                                TensorQuantizer(disabled)
+model.layers.5.self_attn.q_b_proj.output_quantizer                               TensorQuantizer(disabled)
+model.layers.5.self_attn.q_b_proj.weight_quantizer                               TensorQuantizer(disabled)
+model.layers.5.self_attn.kv_a_proj_with_mqa.input_quantizer                      TensorQuantizer(disabled)
+model.layers.5.self_attn.kv_a_proj_with_mqa.output_quantizer                     TensorQuantizer(disabled)
+model.layers.5.self_attn.kv_a_proj_with_mqa.weight_quantizer                     TensorQuantizer(disabled)
+model.layers.5.self_attn.kv_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.5.self_attn.kv_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.5.self_attn.kv_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.5.self_attn.o_proj.input_quantizer                                  TensorQuantizer(disabled)
+model.layers.5.self_attn.o_proj.output_quantizer                                 TensorQuantizer(disabled)
+model.layers.5.self_attn.o_proj.weight_quantizer                                 TensorQuantizer(disabled)
+model.layers.5.self_attn.q_bmm_quantizer                                         TensorQuantizer(disabled)
+model.layers.5.self_attn.k_bmm_quantizer                                         TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.5.self_attn.v_bmm_quantizer                                         TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.5.self_attn.softmax_quantizer                                       TensorQuantizer(disabled)
+model.layers.5.mlp.experts.gate_up_proj_input_quantizer                          TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.24e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.0                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.1                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.2                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.3                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.28e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.4                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.5                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.6                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.75e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.7                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.8                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.9                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.10                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.11                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.12                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.07e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.13                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.14                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.15                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.16                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.89e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.17                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.18                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.19                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.20                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.21                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.22                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.86e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.23                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.87e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.24                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.85e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.25                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.26                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.27                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.28                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.29                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.87e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.30                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.31                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.32                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.33                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.77e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.34                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.35                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.36                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.37                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.38                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.39                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.84e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.40                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.41                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.42                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.43                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.44                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.89e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.45                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.46                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.75e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.47                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.48                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.49                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.50                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.51                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.52                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.53                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.78e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.54                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.55                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.56                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.57                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.58                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.59                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.60                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.61                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.62                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.76e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.63                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.76e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.64                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.65                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.66                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.67                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.68                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.69                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.70                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.71                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.00e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.72                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.73                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.74                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.87e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.75                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.79e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.76                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.77                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.78                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.79                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.80                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.81                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.99e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.82                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.52e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.83                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.84                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.05e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.85                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.86                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.87                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.88                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.89                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.90                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.91                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.90e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.92                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.66e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.93                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.94                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.95                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.96                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.97                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.98                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.85e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.99                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.100                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.101                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.102                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.103                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.104                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.76e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.105                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.106                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.107                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.108                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.109                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.110                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.111                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.112                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.78e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.113                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.114                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.97e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.115                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.91e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.116                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.117                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.118                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.87e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.119                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.120                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.121                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.86e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.122                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.123                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.124                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.85e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.125                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.32e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.126                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.127                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.18e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.128                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.129                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.130                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.131                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.132                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.133                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.78e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.134                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.135                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.136                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.78e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.137                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.94e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.138                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.139                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.140                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.01e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.141                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.14e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.142                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.143                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.144                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.145                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.146                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.30e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.147                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.63e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.148                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.149                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.150                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.151                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.152                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.153                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.84e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.154                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.155                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.156                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.84e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.157                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.158                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.159                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.160                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.58e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.161                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.97e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.162                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.163                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.82e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.164                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.165                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.166                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.167                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.87e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.168                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.169                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.91e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.170                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.171                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.89e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.172                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.173                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.174                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.175                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.05e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.176                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.177                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.02e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.178                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.01e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.179                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.180                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.82e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.181                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.182                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.183                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.11e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.184                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.00e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.185                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.186                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.187                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.188                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.84e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.189                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.190                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.191                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.192                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.193                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.41e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.194                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.03e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.195                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.87e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.196                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.73e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.197                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.198                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.199                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.200                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.96e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.201                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.202                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.203                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.204                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.205                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.206                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.207                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.97e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.208                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.94e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.209                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.210                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.211                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.212                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.213                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.214                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.215                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.216                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.217                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.218                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.219                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.220                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.221                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.79e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.222                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.223                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.224                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.225                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.84e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.226                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.227                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.228                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.229                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.11e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.230                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.231                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.232                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.233                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.234                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.64e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.235                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.73e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.236                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.237                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.238                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.239                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.240                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.241                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.90e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.242                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.90e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.243                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.244                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.245                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.246                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.247                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.19e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.248                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.249                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.92e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.250                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.251                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.252                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.253                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.254                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.255                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.46e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_input_quantizer                             TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.87e+00 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.0                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.1                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.2                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.3                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.4                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.5                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.80e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.6                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.7                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.8                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.9                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.10                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.11                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.12                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.13                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.14                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.15                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.16                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.17                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.18                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.06e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.19                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.92e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.20                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.21                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.22                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.23                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.68e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.24                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.25                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.26                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.27                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.28                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.46e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.29                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.30                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.31                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.32                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.33                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.34                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.35                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.36                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.37                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.38                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.39                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.40                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.41                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.42                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.43                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.44                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.45                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.46                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.47                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.18e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.48                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.49                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.50                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.51                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.18e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.52                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.26e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.53                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.54                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.55                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.56                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.57                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.58                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.59                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.60                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.61                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.62                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.93e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.63                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.23e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.64                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.65                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.66                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.67                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.68                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.11e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.69                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.70                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.71                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.72                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.73                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.74                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.78e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.75                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.76                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.77                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.41e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.78                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.79                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.80                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.81                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.82                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.87e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.83                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.84                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.03e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.85                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.86                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.87                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.88                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.89                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.90                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.91                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.92                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.44e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.93                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.94                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.95                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.87e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.96                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.97                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.60e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.98                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.99                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.100                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.101                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.102                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.25e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.103                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.104                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.105                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.106                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.107                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.108                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.109                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.110                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.111                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.112                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.11e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.113                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.114                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.71e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.115                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.116                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.117                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.118                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.02e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.119                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.120                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.121                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.122                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.123                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.124                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.125                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.24e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.126                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.127                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.128                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.129                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.130                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.131                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.132                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.133                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.134                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.135                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.136                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.137                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.138                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.77e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.139                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.140                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.54e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.141                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.19e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.142                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.143                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.144                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.145                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.146                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.147                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.68e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.148                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.149                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.150                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.151                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.73e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.152                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.153                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.94e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.154                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.155                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.156                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.157                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.158                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.159                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.160                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.53e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.161                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.162                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.163                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.164                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.165                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.166                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.167                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.56e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.168                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.169                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.170                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.171                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.172                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.173                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.174                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.175                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.70e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.176                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.177                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.178                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.179                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.180                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.181                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.182                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.183                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.184                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.185                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.186                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.187                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.07e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.188                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.189                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.190                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.191                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.192                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.193                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.194                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.195                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.24e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.196                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.197                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.198                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.199                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.77e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.200                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.201                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.202                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.203                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.204                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.205                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.206                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.207                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.208                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.92e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.209                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.210                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.211                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.212                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.213                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.214                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.215                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.216                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.217                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.218                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.219                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.220                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.221                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.222                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.89e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.223                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.224                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.225                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.226                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.227                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.10e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.228                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.229                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.230                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.231                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.232                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.05e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.233                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.234                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.83e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.235                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.236                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.57e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.237                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.238                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.06e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.239                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.240                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.241                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.26e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.242                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.85e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.243                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.244                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.245                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.246                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.29e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.247                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.89e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.248                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.249                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.250                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.251                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.252                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.253                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.254                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.experts.down_proj_weight_quantizers.255                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.45e-01 calibrator=MaxCalibrator quant)
+model.layers.5.mlp.shared_experts.gate_proj.input_quantizer                      TensorQuantizer(disabled)
+model.layers.5.mlp.shared_experts.gate_proj.output_quantizer                     TensorQuantizer(disabled)
+model.layers.5.mlp.shared_experts.gate_proj.weight_quantizer                     TensorQuantizer(disabled)
+model.layers.5.mlp.shared_experts.up_proj.input_quantizer                        TensorQuantizer(disabled)
+model.layers.5.mlp.shared_experts.up_proj.output_quantizer                       TensorQuantizer(disabled)
+model.layers.5.mlp.shared_experts.up_proj.weight_quantizer                       TensorQuantizer(disabled)
+model.layers.5.mlp.shared_experts.down_proj.input_quantizer                      TensorQuantizer(disabled)
+model.layers.5.mlp.shared_experts.down_proj.output_quantizer                     TensorQuantizer(disabled)
+model.layers.5.mlp.shared_experts.down_proj.weight_quantizer                     TensorQuantizer(disabled)
+model.layers.6.self_attn.q_a_proj.input_quantizer                                TensorQuantizer(disabled)
+model.layers.6.self_attn.q_a_proj.output_quantizer                               TensorQuantizer(disabled)
+model.layers.6.self_attn.q_a_proj.weight_quantizer                               TensorQuantizer(disabled)
+model.layers.6.self_attn.q_b_proj.input_quantizer                                TensorQuantizer(disabled)
+model.layers.6.self_attn.q_b_proj.output_quantizer                               TensorQuantizer(disabled)
+model.layers.6.self_attn.q_b_proj.weight_quantizer                               TensorQuantizer(disabled)
+model.layers.6.self_attn.kv_a_proj_with_mqa.input_quantizer                      TensorQuantizer(disabled)
+model.layers.6.self_attn.kv_a_proj_with_mqa.output_quantizer                     TensorQuantizer(disabled)
+model.layers.6.self_attn.kv_a_proj_with_mqa.weight_quantizer                     TensorQuantizer(disabled)
+model.layers.6.self_attn.kv_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.6.self_attn.kv_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.6.self_attn.kv_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.6.self_attn.o_proj.input_quantizer                                  TensorQuantizer(disabled)
+model.layers.6.self_attn.o_proj.output_quantizer                                 TensorQuantizer(disabled)
+model.layers.6.self_attn.o_proj.weight_quantizer                                 TensorQuantizer(disabled)
+model.layers.6.self_attn.indexer.wq_b.input_quantizer                            TensorQuantizer(disabled)
+model.layers.6.self_attn.indexer.wq_b.output_quantizer                           TensorQuantizer(disabled)
+model.layers.6.self_attn.indexer.wq_b.weight_quantizer                           TensorQuantizer(disabled)
+model.layers.6.self_attn.indexer.wk.input_quantizer                              TensorQuantizer(disabled)
+model.layers.6.self_attn.indexer.wk.output_quantizer                             TensorQuantizer(disabled)
+model.layers.6.self_attn.indexer.wk.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.6.self_attn.indexer.k_norm.input_quantizer                          TensorQuantizer(disabled)
+model.layers.6.self_attn.indexer.k_norm.output_quantizer                         TensorQuantizer(disabled)
+model.layers.6.self_attn.indexer.weights_proj.input_quantizer                    TensorQuantizer(disabled)
+model.layers.6.self_attn.indexer.weights_proj.output_quantizer                   TensorQuantizer(disabled)
+model.layers.6.self_attn.indexer.weights_proj.weight_quantizer                   TensorQuantizer(disabled)
+model.layers.6.self_attn.q_bmm_quantizer                                         TensorQuantizer(disabled)
+model.layers.6.self_attn.k_bmm_quantizer                                         TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.6.self_attn.v_bmm_quantizer                                         TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.6.self_attn.softmax_quantizer                                       TensorQuantizer(disabled)
+model.layers.6.mlp.experts.gate_up_proj_input_quantizer                          TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.80e+00 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.0                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.1                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.2                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.3                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.4                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.5                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.6                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.7                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.8                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.9                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.90e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.10                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.11                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.12                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.13                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.32e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.14                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.15                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.16                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.17                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.18                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.19                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.20                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.21                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.11e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.22                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.23                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.24                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.25                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.26                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.27                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.28                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.29                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.30                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.31                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.32                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.33                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.34                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.35                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.49e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.36                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.37                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.80e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.38                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.39                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.40                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.41                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.42                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.43                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.44                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.45                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.46                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.47                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.48                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.49                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.50                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.51                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.52                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.53                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.24e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.54                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.73e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.55                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.56                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.57                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.58                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.59                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.60                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.76e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.61                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.62                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.82e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.63                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.64                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.82e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.65                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.66                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.67                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.68                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.69                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.70                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.71                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.72                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.73                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.74                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.75                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.76                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.77                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.78                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.79                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.80                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.81                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.82                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.83                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.84                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.85                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.86                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.87                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.38e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.88                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.89                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.90                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.91                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.92                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.93                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.94                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.95                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.96                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.97                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.98                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.99                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.100                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.101                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.28e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.102                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.103                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.104                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.105                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.106                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.22e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.107                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.108                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.109                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.110                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.111                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.112                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.113                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.114                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.115                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.116                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.117                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.118                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.119                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.67e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.120                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.75e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.121                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.122                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.123                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.124                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.125                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.126                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.07e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.127                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.128                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.129                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.130                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.131                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.132                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.133                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.07e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.134                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.135                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.136                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.137                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.138                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.139                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.45e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.140                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.141                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.142                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.143                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.144                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.145                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.146                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.147                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.148                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.05e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.149                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.150                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.151                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.152                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.153                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.33e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.154                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.155                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.156                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.157                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.158                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.159                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.160                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.161                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.162                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.163                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.164                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.165                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.166                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.77e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.167                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.168                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.169                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.170                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.85e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.171                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.172                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.173                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.174                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.175                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.176                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.91e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.177                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.178                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.179                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.180                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.181                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.182                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.183                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.184                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.185                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.186                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.187                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.188                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.189                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.190                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.191                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.192                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.193                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.194                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.195                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.196                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.197                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.198                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.199                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.200                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.201                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.202                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.203                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.204                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.205                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.206                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.207                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.87e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.208                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.209                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.210                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.211                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.212                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.75e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.213                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.214                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.215                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.216                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.217                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.218                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.86e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.219                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.220                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.221                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.222                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.223                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.224                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.225                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.226                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.227                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.228                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.229                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.230                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.231                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.75e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.232                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.233                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.234                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.69e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.235                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.236                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.237                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.238                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.239                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.240                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.60e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.241                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.242                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.243                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.244                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.245                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.246                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.247                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.248                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.249                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.94e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.250                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.251                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.252                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.253                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.254                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.18e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.255                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_input_quantizer                             TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e+00 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.0                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.1                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.2                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.95e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.3                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.4                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.5                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.6                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.7                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.8                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.9                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.10                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.11                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.80e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.12                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.13                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.27e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.14                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.15                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.16                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.17                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.18                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.19                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.20                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.82e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.21                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.97e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.22                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.23                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.24                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.25                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.26                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.27                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.04e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.28                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.29                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.30                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.31                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.32                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.33                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.34                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.35                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.36                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.37                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.38                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.39                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.40                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.41                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.42                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.43                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.44                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.45                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.46                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.47                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.48                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.49                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.50                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.51                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.52                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.53                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.54                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.55                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.56                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.57                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.58                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.59                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.60                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.61                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.62                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.52e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.63                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.64                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.26e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.65                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.66                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.67                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.68                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.69                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.70                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.71                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.72                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.73                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.74                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.75                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.76                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.77                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.78                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.79                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.80                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.81                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.82                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.56e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.83                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.84                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.85                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.86                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.87                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.42e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.88                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.89                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.90                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.91                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.92                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.93                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.94                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.95                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.96                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.97                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.98                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.99                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.100                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.101                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.45e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.102                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.103                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.104                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.105                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.106                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.68e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.107                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.108                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.109                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.110                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.111                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.112                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.113                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.114                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.115                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.116                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.117                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.118                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.90e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.119                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.93e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.120                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.38e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.121                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.122                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.123                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.124                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.125                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.126                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.86e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.127                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.128                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.129                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.130                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.131                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.132                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.133                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.134                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.135                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.136                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.137                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.138                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.139                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.140                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.141                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.142                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.143                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.144                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.145                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.146                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.147                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.148                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.29e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.149                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.150                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.151                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.54e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.152                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.153                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.154                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.155                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.22e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.156                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.157                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.158                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.159                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.160                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.37e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.161                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.162                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.163                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.164                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.165                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.166                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.167                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.168                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.169                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.170                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.171                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.172                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.173                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.01e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.174                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.175                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.176                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.177                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.178                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.179                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.180                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.181                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.182                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.183                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.184                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.185                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.22e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.186                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.187                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.188                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.189                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.190                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.191                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.95e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.192                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.193                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.194                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.195                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.85e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.196                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.197                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.79e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.198                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.199                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.200                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.201                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.202                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.203                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.204                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.205                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.82e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.206                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.207                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.62e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.208                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.209                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.210                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.211                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.75e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.212                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.213                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.214                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.215                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.216                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.06e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.217                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.218                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.219                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.220                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.221                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.222                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.223                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.224                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.225                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.226                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.227                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.228                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.229                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.230                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.231                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.232                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.233                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.234                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.56e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.235                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.236                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.237                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.238                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.239                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.240                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.85e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.241                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.242                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.243                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.244                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.245                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.246                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.247                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.248                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.249                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.87e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.250                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.251                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.252                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.253                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.254                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.81e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.experts.down_proj_weight_quantizers.255                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.6.mlp.shared_experts.gate_proj.input_quantizer                      TensorQuantizer(disabled)
+model.layers.6.mlp.shared_experts.gate_proj.output_quantizer                     TensorQuantizer(disabled)
+model.layers.6.mlp.shared_experts.gate_proj.weight_quantizer                     TensorQuantizer(disabled)
+model.layers.6.mlp.shared_experts.up_proj.input_quantizer                        TensorQuantizer(disabled)
+model.layers.6.mlp.shared_experts.up_proj.output_quantizer                       TensorQuantizer(disabled)
+model.layers.6.mlp.shared_experts.up_proj.weight_quantizer                       TensorQuantizer(disabled)
+model.layers.6.mlp.shared_experts.down_proj.input_quantizer                      TensorQuantizer(disabled)
+model.layers.6.mlp.shared_experts.down_proj.output_quantizer                     TensorQuantizer(disabled)
+model.layers.6.mlp.shared_experts.down_proj.weight_quantizer                     TensorQuantizer(disabled)
+model.layers.7.self_attn.q_a_proj.input_quantizer                                TensorQuantizer(disabled)
+model.layers.7.self_attn.q_a_proj.output_quantizer                               TensorQuantizer(disabled)
+model.layers.7.self_attn.q_a_proj.weight_quantizer                               TensorQuantizer(disabled)
+model.layers.7.self_attn.q_b_proj.input_quantizer                                TensorQuantizer(disabled)
+model.layers.7.self_attn.q_b_proj.output_quantizer                               TensorQuantizer(disabled)
+model.layers.7.self_attn.q_b_proj.weight_quantizer                               TensorQuantizer(disabled)
+model.layers.7.self_attn.kv_a_proj_with_mqa.input_quantizer                      TensorQuantizer(disabled)
+model.layers.7.self_attn.kv_a_proj_with_mqa.output_quantizer                     TensorQuantizer(disabled)
+model.layers.7.self_attn.kv_a_proj_with_mqa.weight_quantizer                     TensorQuantizer(disabled)
+model.layers.7.self_attn.kv_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.7.self_attn.kv_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.7.self_attn.kv_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.7.self_attn.o_proj.input_quantizer                                  TensorQuantizer(disabled)
+model.layers.7.self_attn.o_proj.output_quantizer                                 TensorQuantizer(disabled)
+model.layers.7.self_attn.o_proj.weight_quantizer                                 TensorQuantizer(disabled)
+model.layers.7.self_attn.q_bmm_quantizer                                         TensorQuantizer(disabled)
+model.layers.7.self_attn.k_bmm_quantizer                                         TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.7.self_attn.v_bmm_quantizer                                         TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.7.self_attn.softmax_quantizer                                       TensorQuantizer(disabled)
+model.layers.7.mlp.experts.gate_up_proj_input_quantizer                          TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e+00 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.0                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.1                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.10e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.2                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.3                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.4                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.00e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.5                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.6                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.7                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.8                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.9                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.10                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.11                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.12                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.13                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.14                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.15                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.16                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.17                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.18                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.19                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.20                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.21                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.22                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.23                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.24                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.25                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.26                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.27                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.78e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.28                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.29                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.30                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.31                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.32                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.33                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.34                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.35                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.36                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.37                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.38                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.39                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.40                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.41                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.42                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.43                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.44                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.45                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.46                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.47                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.48                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.49                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.50                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.51                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.84e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.52                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.53                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.54                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.55                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.56                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.57                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.58                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.59                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.60                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.61                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.62                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.63                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.64                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.65                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.66                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.67                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.68                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.69                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.70                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.71                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.72                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.73                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.74                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.75                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.76                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.77                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.78                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.79                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.80                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.81                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.82                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.83                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.84                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.85                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.86                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.87                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.88                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.89                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.90                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.91                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.92                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.93                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.94                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.95                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.96                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.97                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.98                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.99                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.100                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.101                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.102                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.103                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.104                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.105                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.106                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.107                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.108                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.109                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.110                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.111                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.112                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.113                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.114                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.115                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.116                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.117                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.118                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.119                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.120                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.121                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.122                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.123                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.124                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.125                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.126                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.127                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.128                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.129                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.130                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.131                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.132                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.133                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.134                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.135                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.136                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.137                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.138                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.139                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.140                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.141                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.142                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.143                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.144                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.145                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.146                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.147                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.148                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.149                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.150                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.151                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.152                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.153                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.154                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.155                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.156                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.157                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.158                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.159                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.160                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.161                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.162                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.163                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.164                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.165                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.166                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.167                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.168                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.169                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.170                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.171                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.172                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.173                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.174                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.175                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.176                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.177                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.178                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.179                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.180                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.181                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.182                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.183                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.184                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.185                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.186                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.187                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.188                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.189                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.190                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.191                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.192                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.193                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.194                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.195                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.196                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.197                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.198                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.48e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.199                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.200                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.201                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.202                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.12e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.203                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.204                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.205                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.206                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.207                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.208                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.209                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.210                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.211                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.212                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.213                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.214                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.215                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.66e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.216                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.217                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.218                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.219                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.220                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.221                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.222                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.223                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.224                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.225                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.226                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.227                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.228                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.229                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.230                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.231                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.232                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.233                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.234                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.235                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.236                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.237                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.238                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.239                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.240                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.241                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.242                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.243                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.244                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.245                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.246                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.247                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.248                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.249                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.250                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.251                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.252                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.253                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.254                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.255                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_input_quantizer                             TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.16e+00 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.0                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.1                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.20e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.2                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.3                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.4                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.95e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.5                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.83e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.6                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.7                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.8                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.9                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.10                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.11                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.12                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.13                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.14                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.06e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.15                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.16                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.17                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.18                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.19                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.20                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.21                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.96e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.22                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.23                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.24                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.25                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.26                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.27                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.28                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.29                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.30                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.31                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.32                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.33                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.34                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.35                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.36                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.37                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.38                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.39                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.00e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.40                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.41                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.42                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.43                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.44                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.45                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.46                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.47                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.48                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.49                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.50                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.51                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e+00 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.52                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.53                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.54                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.55                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.56                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.57                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.58                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.59                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.60                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.61                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.62                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.02e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.63                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.64                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.65                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.66                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.78e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.67                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.68                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.69                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.85e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.70                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.71                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.72                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.73                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.74                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.75                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.76                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.77                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.78                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.88e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.79                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.80                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.81                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.82                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.83                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.84                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.85                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.86                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.87                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.27e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.88                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.89                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.90                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.91                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.92                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.35e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.93                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.94                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.95                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.96                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.97                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.98                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.99                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.100                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.101                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.102                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.103                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.104                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.105                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.106                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.107                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.108                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.109                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.110                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.47e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.111                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.112                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.98e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.113                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.114                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.115                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.116                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.117                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.12e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.118                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.119                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.120                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.121                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.122                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.123                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.09e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.124                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.125                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.126                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.127                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.24e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.128                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.129                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.130                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.131                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.132                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.133                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.134                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.135                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.136                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.137                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.138                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.139                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.140                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.141                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.142                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.143                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.144                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.145                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.146                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.68e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.147                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.148                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.149                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.150                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.151                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.152                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.153                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.154                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.155                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.92e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.156                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.157                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.158                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.159                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.160                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.161                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.162                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.163                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.164                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.165                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.166                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.167                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.168                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.169                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.170                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.171                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.172                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.173                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.174                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.175                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.176                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.177                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.178                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.179                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.180                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.181                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.182                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.183                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.184                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.185                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.76e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.186                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.187                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.188                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.189                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.190                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.191                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.192                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.193                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.194                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.195                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.196                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.197                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.198                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.79e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.199                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.27e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.200                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.201                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.202                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.95e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.203                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.204                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.205                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.206                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.207                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.208                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.79e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.209                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.210                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.211                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.212                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.75e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.213                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.214                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.215                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.21e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.216                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.217                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.218                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.219                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.220                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.221                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.222                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.223                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.224                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.75e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.225                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.226                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.227                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.228                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.229                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.230                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.231                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.232                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.233                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.234                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.235                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.236                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.237                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.238                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.239                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.240                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.241                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.242                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.243                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.244                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.245                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.246                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.247                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.248                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.249                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.250                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.251                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.252                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.253                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.254                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.82e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.experts.down_proj_weight_quantizers.255                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.7.mlp.shared_experts.gate_proj.input_quantizer                      TensorQuantizer(disabled)
+model.layers.7.mlp.shared_experts.gate_proj.output_quantizer                     TensorQuantizer(disabled)
+model.layers.7.mlp.shared_experts.gate_proj.weight_quantizer                     TensorQuantizer(disabled)
+model.layers.7.mlp.shared_experts.up_proj.input_quantizer                        TensorQuantizer(disabled)
+model.layers.7.mlp.shared_experts.up_proj.output_quantizer                       TensorQuantizer(disabled)
+model.layers.7.mlp.shared_experts.up_proj.weight_quantizer                       TensorQuantizer(disabled)
+model.layers.7.mlp.shared_experts.down_proj.input_quantizer                      TensorQuantizer(disabled)
+model.layers.7.mlp.shared_experts.down_proj.output_quantizer                     TensorQuantizer(disabled)
+model.layers.7.mlp.shared_experts.down_proj.weight_quantizer                     TensorQuantizer(disabled)
+model.layers.8.self_attn.q_a_proj.input_quantizer                                TensorQuantizer(disabled)
+model.layers.8.self_attn.q_a_proj.output_quantizer                               TensorQuantizer(disabled)
+model.layers.8.self_attn.q_a_proj.weight_quantizer                               TensorQuantizer(disabled)
+model.layers.8.self_attn.q_b_proj.input_quantizer                                TensorQuantizer(disabled)
+model.layers.8.self_attn.q_b_proj.output_quantizer                               TensorQuantizer(disabled)
+model.layers.8.self_attn.q_b_proj.weight_quantizer                               TensorQuantizer(disabled)
+model.layers.8.self_attn.kv_a_proj_with_mqa.input_quantizer                      TensorQuantizer(disabled)
+model.layers.8.self_attn.kv_a_proj_with_mqa.output_quantizer                     TensorQuantizer(disabled)
+model.layers.8.self_attn.kv_a_proj_with_mqa.weight_quantizer                     TensorQuantizer(disabled)
+model.layers.8.self_attn.kv_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.8.self_attn.kv_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.8.self_attn.kv_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.8.self_attn.o_proj.input_quantizer                                  TensorQuantizer(disabled)
+model.layers.8.self_attn.o_proj.output_quantizer                                 TensorQuantizer(disabled)
+model.layers.8.self_attn.o_proj.weight_quantizer                                 TensorQuantizer(disabled)
+model.layers.8.self_attn.q_bmm_quantizer                                         TensorQuantizer(disabled)
+model.layers.8.self_attn.k_bmm_quantizer                                         TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.8.self_attn.v_bmm_quantizer                                         TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.8.self_attn.softmax_quantizer                                       TensorQuantizer(disabled)
+model.layers.8.mlp.experts.gate_up_proj_input_quantizer                          TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e+00 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.0                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.1                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.2                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.75e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.3                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.4                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.91e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.5                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.24e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.6                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.7                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.8                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.9                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.04e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.10                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.11                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.12                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.13                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.05e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.14                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.15                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.77e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.16                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.17                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.18                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.95e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.19                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.99e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.20                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.21                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.22                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.23e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.23                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.24                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.25                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.26                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.27                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.96e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.28                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.29                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.30                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.31                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.32                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.33                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.34                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.21e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.35                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.36                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.37                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.38                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.39                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.40                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.70e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.41                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.09e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.42                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.43                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.22e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.44                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.45                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.29e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.46                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.47                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.48                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.15e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.49                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.91e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.50                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.51                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.24e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.52                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.53                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.79e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.54                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.55                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.56                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.57                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.52e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.58                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.59                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.60                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.61                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.62                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.63                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.64                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.11e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.65                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.66                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.67                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.28e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.68                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.69                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.70                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.71                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.72                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.06e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.73                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.74                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.09e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.75                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.76                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.77                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.78                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.09e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.79                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.80                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.81                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.82                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.83                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.84                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.85                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.86                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.02e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.87                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.88                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.64e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.89                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.90                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.91                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.20e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.92                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.93                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.94                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.95                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.96                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.97                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.91e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.98                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.99                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.100                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.101                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.08e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.102                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.17e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.103                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.104                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.98e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.105                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.106                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.107                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.108                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.109                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.110                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.111                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.112                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.113                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.19e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.114                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.115                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.116                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.117                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.71e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.118                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.119                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.120                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.121                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.122                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.94e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.123                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.124                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.125                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.126                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.127                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.128                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.129                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.130                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.25e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.131                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.18e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.132                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.133                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.134                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.135                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.77e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.136                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.15e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.137                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.138                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.139                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.140                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.88e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.141                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.142                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.143                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.144                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.145                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.41e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.146                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.147                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.148                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.149                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.150                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.151                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.152                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.07e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.153                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.154                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.155                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.22e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.156                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.157                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.158                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.26e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.159                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.77e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.160                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.161                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.162                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.163                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.164                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.165                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.166                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.167                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.168                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.169                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.170                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.171                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.172                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.44e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.173                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.22e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.174                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.175                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.176                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.177                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.26e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.178                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.179                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.180                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.181                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.06e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.182                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.07e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.183                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.184                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.64e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.185                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.186                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.14e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.187                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.188                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.01e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.189                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.190                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.191                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.192                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.193                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.194                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.89e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.195                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.85e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.196                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.197                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.198                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.199                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.49e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.200                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.29e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.201                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.202                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.203                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.204                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.205                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.80e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.206                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.76e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.207                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.78e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.208                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.209                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.08e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.210                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.211                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.24e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.212                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.213                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.214                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.215                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.90e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.216                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.83e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.217                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.94e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.218                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.219                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.09e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.220                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.13e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.221                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.222                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.223                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.24e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.224                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.225                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.226                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.62e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.227                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.96e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.228                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.229                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.96e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.230                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.38e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.231                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.232                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.233                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.234                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.22e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.235                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.236                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.237                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.73e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.238                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.239                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.240                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.241                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.242                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.243                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.21e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.244                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.245                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.246                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.247                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.248                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.249                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.250                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.21e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.251                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.252                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.253                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.254                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.255                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_input_quantizer                             TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.60e+01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.0                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.1                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.2                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.3                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.4                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.70e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.5                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.25e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.6                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.7                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.8                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.9                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.10                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.11                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.12                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.13                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.14                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.15                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.16                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.17                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.18                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.19                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.20                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.21                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.22                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.23                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.24                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.25                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.26                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.27                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.73e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.28                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.29                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.30                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.31                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.32                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.33                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.34                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.35                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.36                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.37                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.38                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.39                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.40                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.41                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.42                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.43                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.44                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.45                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.46                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.47                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.48                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.49                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.50                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.51                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.52                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.53                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.54                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.55                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.73e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.56                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.57                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.58                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.59                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.60                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.61                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.62                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.63                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.64                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.65                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.66                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.67                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.68                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.69                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.70                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.71                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.72                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.73                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.74                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.75                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.76                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.77                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.78                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.79                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.80                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.81                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.82                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.83                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.84                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.85                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.86                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.36e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.87                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.88                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.89                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.90                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.91                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.92                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.93                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.94                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.95                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.96                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.97                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.98                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.99                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.100                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.101                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.102                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.25e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.103                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.104                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.105                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.106                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.91e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.107                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.80e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.108                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.109                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.110                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.111                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.112                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.113                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.114                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.115                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.116                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.117                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.118                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.119                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.120                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.121                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.122                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.123                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.124                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.125                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.126                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.127                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.128                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.129                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.130                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.131                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.132                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.133                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.134                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.135                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.136                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.137                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.138                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.139                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.140                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.02e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.141                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.142                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.143                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.144                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.145                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.146                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.147                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.148                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.149                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.150                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.151                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.20e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.152                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.153                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.154                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.98e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.155                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.52e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.156                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.157                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.158                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.159                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.160                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.161                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.162                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.163                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.164                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.165                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.13e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.166                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.167                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.168                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.169                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.170                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.171                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.172                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.173                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.95e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.174                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.175                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.76e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.176                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.177                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.178                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.179                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.180                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.181                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.182                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.183                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.184                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.185                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.186                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.187                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.188                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.189                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.190                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.191                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.192                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.193                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.194                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.195                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.196                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.197                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.198                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.199                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.200                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.201                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.202                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.203                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.204                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.205                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.206                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.70e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.207                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.208                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.209                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.210                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.211                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.18e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.212                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.213                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.214                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.215                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.216                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.217                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.218                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.219                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.220                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.221                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.222                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.223                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.224                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.225                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.13e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.226                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.227                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.228                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.229                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.230                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.231                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.232                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.233                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.234                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.235                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.236                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.237                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.238                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.239                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.240                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.241                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.242                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.243                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.78e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.244                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.245                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.246                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.247                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.248                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.249                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.250                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.251                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.252                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.253                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.254                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.92e-01 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.experts.down_proj_weight_quantizers.255                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.8.mlp.shared_experts.gate_proj.input_quantizer                      TensorQuantizer(disabled)
+model.layers.8.mlp.shared_experts.gate_proj.output_quantizer                     TensorQuantizer(disabled)
+model.layers.8.mlp.shared_experts.gate_proj.weight_quantizer                     TensorQuantizer(disabled)
+model.layers.8.mlp.shared_experts.up_proj.input_quantizer                        TensorQuantizer(disabled)
+model.layers.8.mlp.shared_experts.up_proj.output_quantizer                       TensorQuantizer(disabled)
+model.layers.8.mlp.shared_experts.up_proj.weight_quantizer                       TensorQuantizer(disabled)
+model.layers.8.mlp.shared_experts.down_proj.input_quantizer                      TensorQuantizer(disabled)
+model.layers.8.mlp.shared_experts.down_proj.output_quantizer                     TensorQuantizer(disabled)
+model.layers.8.mlp.shared_experts.down_proj.weight_quantizer                     TensorQuantizer(disabled)
+model.layers.9.self_attn.q_a_proj.input_quantizer                                TensorQuantizer(disabled)
+model.layers.9.self_attn.q_a_proj.output_quantizer                               TensorQuantizer(disabled)
+model.layers.9.self_attn.q_a_proj.weight_quantizer                               TensorQuantizer(disabled)
+model.layers.9.self_attn.q_b_proj.input_quantizer                                TensorQuantizer(disabled)
+model.layers.9.self_attn.q_b_proj.output_quantizer                               TensorQuantizer(disabled)
+model.layers.9.self_attn.q_b_proj.weight_quantizer                               TensorQuantizer(disabled)
+model.layers.9.self_attn.kv_a_proj_with_mqa.input_quantizer                      TensorQuantizer(disabled)
+model.layers.9.self_attn.kv_a_proj_with_mqa.output_quantizer                     TensorQuantizer(disabled)
+model.layers.9.self_attn.kv_a_proj_with_mqa.weight_quantizer                     TensorQuantizer(disabled)
+model.layers.9.self_attn.kv_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.9.self_attn.kv_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.9.self_attn.kv_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.9.self_attn.o_proj.input_quantizer                                  TensorQuantizer(disabled)
+model.layers.9.self_attn.o_proj.output_quantizer                                 TensorQuantizer(disabled)
+model.layers.9.self_attn.o_proj.weight_quantizer                                 TensorQuantizer(disabled)
+model.layers.9.self_attn.q_bmm_quantizer                                         TensorQuantizer(disabled)
+model.layers.9.self_attn.k_bmm_quantizer                                         TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.9.self_attn.v_bmm_quantizer                                         TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.9.self_attn.softmax_quantizer                                       TensorQuantizer(disabled)
+model.layers.9.mlp.experts.gate_up_proj_input_quantizer                          TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.03e+00 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.0                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.1                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.2                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.3                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.4                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.5                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.6                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.7                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.8                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.9                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.10                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.11                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.12                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.13                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.14                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.15                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.28e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.16                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.17                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.18                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.19                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.30e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.20                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.21                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.22                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.23                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.24                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.79e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.25                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.26                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.27                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.28                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.29                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.30                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.31                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.32                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.33                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.34                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.35                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.36                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.37                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.38                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.39                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.40                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.41                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.42                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.43                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.44                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.45                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.46                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.47                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.48                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.49                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.84e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.50                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.51                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.52                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.53                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.54                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.55                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.56                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.57                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.58                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.59                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.60                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.61                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.62                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.63                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.64                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.65                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.66                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.46e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.67                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.68                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.69                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.70                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.71                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.72                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.73                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.99e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.74                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.75                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.76                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.77                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.78                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.79                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.80                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.81                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.82                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.83                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.84                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.01e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.85                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.86                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.87                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.88                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.89                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.90                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.91                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.92                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.93                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.94                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.95                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.96                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.97                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.98                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.99                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.100                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.12e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.101                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.102                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.103                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.104                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.105                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.106                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.107                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.108                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.109                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.110                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.111                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.112                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.113                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.114                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.115                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.116                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.117                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.118                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.119                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.120                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.121                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.122                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.123                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.124                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.125                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.126                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.127                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.128                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.129                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.130                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.131                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.41e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.132                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.133                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.134                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.135                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.136                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.137                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.138                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.139                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.140                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.141                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.142                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.143                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.144                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.145                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.146                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.147                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.148                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.149                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.150                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.151                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.08e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.152                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.153                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.154                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.155                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.156                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.157                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.158                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.159                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.160                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.161                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.162                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.163                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.164                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.165                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.166                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.167                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.168                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.169                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.170                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.36e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.171                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.172                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.173                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.174                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.175                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.176                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.177                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.56e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.178                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.179                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.180                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.181                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.182                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.183                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.184                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.185                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.60e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.186                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.187                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.188                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.189                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.190                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.191                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.192                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.193                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.194                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.195                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.196                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.197                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.198                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.199                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.200                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.201                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.202                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.203                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.204                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.205                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.206                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.207                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.208                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.209                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.210                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.92e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.211                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.212                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.213                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.214                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.215                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.216                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.217                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.218                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.219                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.220                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.221                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.222                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.223                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.224                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.225                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.226                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.227                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.228                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.229                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.230                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.231                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.232                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.233                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.234                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.235                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.236                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.81e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.237                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.238                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.239                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.240                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.241                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.242                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.243                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.244                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.245                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.246                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.247                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.248                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.249                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.250                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.251                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.252                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.253                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.254                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.255                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_input_quantizer                             TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.50e+00 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.0                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.1                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.2                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.90e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.3                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.4                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.52e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.5                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.6                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.7                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.8                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.9                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.10                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.11                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.12                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.13                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.14                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.15                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.77e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.16                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.17                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.18                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.19                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.16e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.20                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.21                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.22                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.23                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.24                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.29e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.25                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.26                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.27                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.28                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.29                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.30                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.31                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.32                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.33                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.34                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.35                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.36                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.37                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.38                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.39                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.40                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.96e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.41                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.42                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.43                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.44                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.45                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.46                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.47                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.48                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.49                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.50                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.51                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.52                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.53                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.54                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.55                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.56                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.57                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.58                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.59                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.60                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.61                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.62                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.63                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.64                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.65                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.66                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.67                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.68                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.69                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.70                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.71                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.72                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.73                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.74                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.75                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.43e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.76                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.77                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.78                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.79                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.80                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.81                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.82                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.83                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.84                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.85                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.86                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.87                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.88                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.89                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.90                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.91                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.92                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.23e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.93                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.94                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.95                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.96                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.97                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.98                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.99                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.100                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.101                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.102                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.103                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.104                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.105                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.106                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.107                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.108                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.109                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.110                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.111                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.112                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.113                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.114                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.115                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.116                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.117                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.118                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.119                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.120                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.121                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.122                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.123                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.124                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.125                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.126                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.127                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.128                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.129                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.130                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.131                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.95e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.132                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.133                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.134                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.135                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.136                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.137                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.138                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.139                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.140                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.82e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.141                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.142                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.143                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.144                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.145                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.146                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.147                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.148                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.149                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.150                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.151                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.93e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.152                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.153                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.154                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.155                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.156                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.157                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.158                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.159                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.160                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.161                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.162                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.163                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.164                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.14e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.165                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.166                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.167                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.168                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.169                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.170                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.171                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.172                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.173                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.174                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.175                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.176                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.177                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.178                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.179                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.180                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.181                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.182                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.183                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.184                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.185                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.186                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.187                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.188                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.189                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.190                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.191                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.192                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.193                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.194                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.195                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.196                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.197                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.198                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.199                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.200                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.201                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.202                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.203                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.15e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.204                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.205                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.206                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.207                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.208                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.209                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.210                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.211                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.212                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.213                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.214                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.215                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.216                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.217                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.218                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.219                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.220                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.79e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.221                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.222                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.223                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.224                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.225                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.226                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.227                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.228                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.229                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.230                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.231                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.232                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.233                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.234                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.235                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.236                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.237                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.238                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.239                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.240                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.241                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.242                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.243                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.83e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.244                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.245                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.246                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.247                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.248                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.249                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.250                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.251                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.252                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.253                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.254                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.experts.down_proj_weight_quantizers.255                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.9.mlp.shared_experts.gate_proj.input_quantizer                      TensorQuantizer(disabled)
+model.layers.9.mlp.shared_experts.gate_proj.output_quantizer                     TensorQuantizer(disabled)
+model.layers.9.mlp.shared_experts.gate_proj.weight_quantizer                     TensorQuantizer(disabled)
+model.layers.9.mlp.shared_experts.up_proj.input_quantizer                        TensorQuantizer(disabled)
+model.layers.9.mlp.shared_experts.up_proj.output_quantizer                       TensorQuantizer(disabled)
+model.layers.9.mlp.shared_experts.up_proj.weight_quantizer                       TensorQuantizer(disabled)
+model.layers.9.mlp.shared_experts.down_proj.input_quantizer                      TensorQuantizer(disabled)
+model.layers.9.mlp.shared_experts.down_proj.output_quantizer                     TensorQuantizer(disabled)
+model.layers.9.mlp.shared_experts.down_proj.weight_quantizer                     TensorQuantizer(disabled)
+model.layers.10.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.10.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.10.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.10.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.10.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.10.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.10.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.10.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.10.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.10.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.10.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.10.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.10.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.10.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.10.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.10.self_attn.indexer.wq_b.input_quantizer                           TensorQuantizer(disabled)
+model.layers.10.self_attn.indexer.wq_b.output_quantizer                          TensorQuantizer(disabled)
+model.layers.10.self_attn.indexer.wq_b.weight_quantizer                          TensorQuantizer(disabled)
+model.layers.10.self_attn.indexer.wk.input_quantizer                             TensorQuantizer(disabled)
+model.layers.10.self_attn.indexer.wk.output_quantizer                            TensorQuantizer(disabled)
+model.layers.10.self_attn.indexer.wk.weight_quantizer                            TensorQuantizer(disabled)
+model.layers.10.self_attn.indexer.k_norm.input_quantizer                         TensorQuantizer(disabled)
+model.layers.10.self_attn.indexer.k_norm.output_quantizer                        TensorQuantizer(disabled)
+model.layers.10.self_attn.indexer.weights_proj.input_quantizer                   TensorQuantizer(disabled)
+model.layers.10.self_attn.indexer.weights_proj.output_quantizer                  TensorQuantizer(disabled)
+model.layers.10.self_attn.indexer.weights_proj.weight_quantizer                  TensorQuantizer(disabled)
+model.layers.10.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.10.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.10.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.10.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.10.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.02e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.19e+00 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.41e+00 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.98e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.02e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.84e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.62e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.78e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.47e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.32e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.02e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.34e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.87e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.11e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.11e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.07e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.77e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.34e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.15e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.10.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.10.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.10.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.10.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.10.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.10.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.10.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.10.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.10.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.11.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.11.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.11.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.11.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.11.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.11.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.11.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.11.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.11.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.11.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.11.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.11.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.11.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.11.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.11.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.11.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.11.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.11.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.11.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.11.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e+00 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.84e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.35e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.96e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.96e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.84e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.79e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.02e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.86e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.97e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.99e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.41e+00 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.04e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.00e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.91e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.29e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.75e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.76e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.18e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.79e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.84e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.27e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.87e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.78e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.75e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.05e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.93e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.93e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.58e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.44e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.11.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.11.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.11.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.11.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.11.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.11.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.11.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.11.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.11.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.12.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.12.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.12.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.12.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.12.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.12.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.12.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.12.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.12.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.12.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.12.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.12.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.12.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.12.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.12.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.12.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.12.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.12.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.12.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.12.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e+00 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.89e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.77e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.87e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.77e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.99e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.12e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.75e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.16e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.06e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.02e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.12e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.76e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.97e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.67e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.83e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.05e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.99e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.01e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.11e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.31e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.08e+00 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.98e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.77e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.55e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.81e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.85e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.54e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.60e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.96e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.25e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.14e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.41e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.97e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.70e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.07e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.66e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.78e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.95e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.56e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.12.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.12.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.12.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.12.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.12.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.12.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.12.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.12.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.12.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.13.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.13.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.13.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.13.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.13.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.13.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.13.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.13.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.13.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.13.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.13.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.13.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.13.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.13.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.13.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.13.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.13.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.13.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.13.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.13.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e+00 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.91e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.73e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.11e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.72e+00 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.00e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.18e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.24e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.97e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.13.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.13.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.13.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.13.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.13.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.13.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.13.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.13.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.13.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.14.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.14.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.14.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.14.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.14.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.14.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.14.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.14.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.14.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.14.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.14.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.14.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.14.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.14.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.14.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.14.self_attn.indexer.wq_b.input_quantizer                           TensorQuantizer(disabled)
+model.layers.14.self_attn.indexer.wq_b.output_quantizer                          TensorQuantizer(disabled)
+model.layers.14.self_attn.indexer.wq_b.weight_quantizer                          TensorQuantizer(disabled)
+model.layers.14.self_attn.indexer.wk.input_quantizer                             TensorQuantizer(disabled)
+model.layers.14.self_attn.indexer.wk.output_quantizer                            TensorQuantizer(disabled)
+model.layers.14.self_attn.indexer.wk.weight_quantizer                            TensorQuantizer(disabled)
+model.layers.14.self_attn.indexer.k_norm.input_quantizer                         TensorQuantizer(disabled)
+model.layers.14.self_attn.indexer.k_norm.output_quantizer                        TensorQuantizer(disabled)
+model.layers.14.self_attn.indexer.weights_proj.input_quantizer                   TensorQuantizer(disabled)
+model.layers.14.self_attn.indexer.weights_proj.output_quantizer                  TensorQuantizer(disabled)
+model.layers.14.self_attn.indexer.weights_proj.weight_quantizer                  TensorQuantizer(disabled)
+model.layers.14.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.14.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.14.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.14.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.14.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e+00 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.09e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.65e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.66e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.01e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.93e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.08e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.26e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.46e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.39e+00 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.22e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.92e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.30e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.76e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.23e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.96e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.96e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.76e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.08e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.90e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.62e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.09e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.80e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.14.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.14.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.14.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.14.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.14.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.14.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.14.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.14.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.14.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.15.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.15.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.15.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.15.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.15.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.15.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.15.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.15.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.15.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.15.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.15.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.15.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.15.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.15.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.15.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.15.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.15.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.15.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.15.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.15.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e+00 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.77e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.71e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.54e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.31e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.98e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.90e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.13e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.75e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.77e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.27e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.76e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.76e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.79e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.98e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.83e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.98e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.90e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.53e+00 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.56e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.00e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.77e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.41e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.75e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.84e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.87e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.82e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.67e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.94e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.23e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.85e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.15.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.15.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.15.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.15.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.15.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.15.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.15.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.15.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.15.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.16.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.16.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.16.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.16.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.16.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.16.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.16.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.16.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.16.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.16.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.16.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.16.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.16.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.16.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.16.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.16.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.16.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.16.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.16.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.16.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e+00 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.35e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.19e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.49e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.01e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.32e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.14e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.09e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.28e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.16e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.09e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.06e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.04e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.23e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.85e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.90e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.87e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.00e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.68e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.64e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.69e+00 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.22e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.48e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.82e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.77e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.12e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.51e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.23e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.89e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.71e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.28e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.01e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.75e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.23e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.16.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.16.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.16.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.16.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.16.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.16.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.16.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.16.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.16.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.17.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.17.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.17.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.17.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.17.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.17.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.17.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.17.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.17.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.17.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.17.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.17.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.17.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.17.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.17.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.17.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.17.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.17.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.17.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.17.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e+00 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.87e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.73e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.06e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.76e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.08e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.91e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.03e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.33e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.75e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.70e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.13e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.30e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.28e+00 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.64e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.83e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.73e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.35e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.30e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.87e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.77e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.83e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.92e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.42e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.89e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.29e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.73e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.16e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.91e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.96e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.29e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.43e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.26e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.52e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.17.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.17.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.17.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.17.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.17.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.17.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.17.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.17.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.17.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.18.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.18.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.18.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.18.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.18.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.18.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.18.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.18.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.18.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.18.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.18.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.18.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.18.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.18.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.18.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.18.self_attn.indexer.wq_b.input_quantizer                           TensorQuantizer(disabled)
+model.layers.18.self_attn.indexer.wq_b.output_quantizer                          TensorQuantizer(disabled)
+model.layers.18.self_attn.indexer.wq_b.weight_quantizer                          TensorQuantizer(disabled)
+model.layers.18.self_attn.indexer.wk.input_quantizer                             TensorQuantizer(disabled)
+model.layers.18.self_attn.indexer.wk.output_quantizer                            TensorQuantizer(disabled)
+model.layers.18.self_attn.indexer.wk.weight_quantizer                            TensorQuantizer(disabled)
+model.layers.18.self_attn.indexer.k_norm.input_quantizer                         TensorQuantizer(disabled)
+model.layers.18.self_attn.indexer.k_norm.output_quantizer                        TensorQuantizer(disabled)
+model.layers.18.self_attn.indexer.weights_proj.input_quantizer                   TensorQuantizer(disabled)
+model.layers.18.self_attn.indexer.weights_proj.output_quantizer                  TensorQuantizer(disabled)
+model.layers.18.self_attn.indexer.weights_proj.weight_quantizer                  TensorQuantizer(disabled)
+model.layers.18.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.18.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.18.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.18.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.18.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e+00 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.01e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.91e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.47e+00 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.84e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.12e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.86e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.19e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.99e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.56e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.18.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.18.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.18.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.18.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.18.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.18.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.18.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.18.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.18.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.19.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.19.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.19.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.19.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.19.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.19.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.19.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.19.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.19.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.19.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.19.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.19.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.19.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.19.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.19.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.19.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.19.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.19.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.19.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.19.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e+00 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.94e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.28e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.87e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.01e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.14e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.03e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.16e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.08e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.71e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.34e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.47e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.99e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.39e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.80e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.78e+00 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.40e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.40e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.33e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.03e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.79e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.30e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.35e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.40e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.03e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.26e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.40e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.78e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.19.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.19.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.19.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.19.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.19.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.19.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.19.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.19.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.19.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.20.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.20.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.20.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.20.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.20.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.20.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.20.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.20.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.20.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.20.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.20.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.20.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.20.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.20.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.20.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.20.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.20.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.20.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.20.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.20.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e+00 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.41e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.90e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.18e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.77e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.95e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.28e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.36e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.34e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.82e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.75e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.44e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.91e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.77e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.14e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.60e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.07e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.20e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.10e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.63e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.11e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.14e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.44e+00 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.18e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.57e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.40e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.75e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.32e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.15e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.68e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.75e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.35e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.40e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.83e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.35e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.71e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.52e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.35e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.40e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.20.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.20.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.20.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.20.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.20.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.20.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.20.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.20.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.20.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.21.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.21.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.21.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.21.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.21.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.21.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.21.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.21.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.21.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.21.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.21.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.21.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.21.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.21.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.21.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.21.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.21.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.21.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.21.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.21.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e+00 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.26e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.02e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.97e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.38e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.95e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.92e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.79e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.14e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.08e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.05e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.54e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.20e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.79e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.02e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.84e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.92e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.34e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.78e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.50e+00 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.33e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.82e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.28e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.22e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.48e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.98e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.45e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.40e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.26e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.41e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.99e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.79e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.80e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.40e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.21.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.21.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.21.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.21.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.21.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.21.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.21.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.21.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.21.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.22.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.22.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.22.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.22.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.22.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.22.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.22.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.22.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.22.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.22.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.22.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.22.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.22.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.22.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.22.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.22.self_attn.indexer.wq_b.input_quantizer                           TensorQuantizer(disabled)
+model.layers.22.self_attn.indexer.wq_b.output_quantizer                          TensorQuantizer(disabled)
+model.layers.22.self_attn.indexer.wq_b.weight_quantizer                          TensorQuantizer(disabled)
+model.layers.22.self_attn.indexer.wk.input_quantizer                             TensorQuantizer(disabled)
+model.layers.22.self_attn.indexer.wk.output_quantizer                            TensorQuantizer(disabled)
+model.layers.22.self_attn.indexer.wk.weight_quantizer                            TensorQuantizer(disabled)
+model.layers.22.self_attn.indexer.k_norm.input_quantizer                         TensorQuantizer(disabled)
+model.layers.22.self_attn.indexer.k_norm.output_quantizer                        TensorQuantizer(disabled)
+model.layers.22.self_attn.indexer.weights_proj.input_quantizer                   TensorQuantizer(disabled)
+model.layers.22.self_attn.indexer.weights_proj.output_quantizer                  TensorQuantizer(disabled)
+model.layers.22.self_attn.indexer.weights_proj.weight_quantizer                  TensorQuantizer(disabled)
+model.layers.22.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.22.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.22.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.22.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.22.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e+00 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.07e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.91e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.06e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.00e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.30e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.31e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.87e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.37e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.36e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.01e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.96e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.83e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.09e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.79e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.75e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.73e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.38e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.84e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.73e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.10e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.20e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.01e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.25e+00 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.40e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.01e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.30e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.30e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.48e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.08e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.02e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.30e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.56e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.35e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.01e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.76e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.82e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.91e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.40e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.94e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.79e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.95e-01 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.22.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.22.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.22.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.22.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.22.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.22.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.22.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.22.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.22.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.23.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.23.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.23.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.23.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.23.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.23.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.23.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.23.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.23.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.23.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.23.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.23.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.23.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.23.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.23.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.23.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.23.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.23.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.23.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.23.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.67e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.79e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.04e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.34e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.59e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.26e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.87e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.08e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.77e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.89e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.94e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.45e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.25e+00 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.46e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.81e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.25e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.16e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.76e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.77e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.84e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.80e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.35e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.87e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.10e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.75e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.94e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.20e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.23.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.23.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.23.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.23.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.23.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.23.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.23.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.23.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.23.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.24.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.24.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.24.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.24.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.24.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.24.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.24.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.24.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.24.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.24.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.24.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.24.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.24.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.24.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.24.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.24.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.24.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.24.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.24.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.24.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.00e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.44e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.07e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.32e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.88e+00 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.30e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.01e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.92e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.40e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.20e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.37e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.89e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.95e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.50e-01 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.24.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.24.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.24.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.24.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.24.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.24.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.24.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.24.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.24.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.25.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.25.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.25.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.25.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.25.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.25.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.25.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.25.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.25.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.25.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.25.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.25.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.25.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.25.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.25.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.25.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.25.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.25.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.25.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.25.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.65e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.87e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.83e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.75e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.02e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.17e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.09e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.93e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.41e+00 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.18e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.52e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.50e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.35e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.30e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.82e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.35e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.20e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.16e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.54e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.58e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.25.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.25.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.25.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.25.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.25.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.25.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.25.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.25.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.25.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.26.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.26.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.26.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.26.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.26.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.26.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.26.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.26.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.26.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.26.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.26.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.26.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.26.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.26.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.26.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.26.self_attn.indexer.wq_b.input_quantizer                           TensorQuantizer(disabled)
+model.layers.26.self_attn.indexer.wq_b.output_quantizer                          TensorQuantizer(disabled)
+model.layers.26.self_attn.indexer.wq_b.weight_quantizer                          TensorQuantizer(disabled)
+model.layers.26.self_attn.indexer.wk.input_quantizer                             TensorQuantizer(disabled)
+model.layers.26.self_attn.indexer.wk.output_quantizer                            TensorQuantizer(disabled)
+model.layers.26.self_attn.indexer.wk.weight_quantizer                            TensorQuantizer(disabled)
+model.layers.26.self_attn.indexer.k_norm.input_quantizer                         TensorQuantizer(disabled)
+model.layers.26.self_attn.indexer.k_norm.output_quantizer                        TensorQuantizer(disabled)
+model.layers.26.self_attn.indexer.weights_proj.input_quantizer                   TensorQuantizer(disabled)
+model.layers.26.self_attn.indexer.weights_proj.output_quantizer                  TensorQuantizer(disabled)
+model.layers.26.self_attn.indexer.weights_proj.weight_quantizer                  TensorQuantizer(disabled)
+model.layers.26.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.26.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.26.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.26.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.26.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.61e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.56e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.87e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.60e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.90e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.05e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.12e+00 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.07e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.40e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.18e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.84e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.35e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.25e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.30e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.26.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.26.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.26.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.26.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.26.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.26.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.26.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.26.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.26.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.27.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.27.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.27.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.27.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.27.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.27.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.27.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.27.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.27.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.27.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.27.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.27.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.27.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.27.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.27.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.27.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.27.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.27.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.27.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.27.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e+00 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.47e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.08e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.81e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.99e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.03e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.16e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.05e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.86e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.91e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.52e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.81e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.76e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.22e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.44e+00 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.23e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.85e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.23e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.62e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.34e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.90e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.21e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.03e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.95e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.01e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.26e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.99e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.71e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.94e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.95e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.27.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.27.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.27.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.27.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.27.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.27.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.27.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.27.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.27.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.28.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.28.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.28.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.28.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.28.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.28.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.28.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.28.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.28.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.28.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.28.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.28.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.28.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.28.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.28.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.28.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.28.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.28.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.28.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.28.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e+00 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.46e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.05e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.95e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.07e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.41e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.02e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.56e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.95e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.38e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.97e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.83e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.83e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.12e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.50e+00 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.35e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.19e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.40e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.40e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.35e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.90e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.85e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.82e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.39e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.73e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.35e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.75e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.35e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.28.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.28.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.28.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.28.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.28.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.28.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.28.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.28.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.28.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.29.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.29.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.29.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.29.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.29.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.29.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.29.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.29.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.29.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.29.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.29.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.29.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.29.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.29.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.29.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.29.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.29.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.29.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.29.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.29.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e+00 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.26e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.89e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.96e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.28e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.78e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.32e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.20e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.42e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.76e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.83e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.14e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.25e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.37e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.22e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.04e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.26e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.94e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.16e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.85e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.34e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.97e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.29e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.01e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e+01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.91e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.15e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.70e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.19e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.63e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.75e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.63e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.48e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.58e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.22e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.99e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.35e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.49e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.99e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.87e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.40e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.77e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.98e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.40e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.89e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.35e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.14e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.79e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.40e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.29.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.29.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.29.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.29.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.29.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.29.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.29.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.29.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.29.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.30.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.30.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.30.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.30.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.30.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.30.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.30.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.30.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.30.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.30.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.30.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.30.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.30.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.30.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.30.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.30.self_attn.indexer.wq_b.input_quantizer                           TensorQuantizer(disabled)
+model.layers.30.self_attn.indexer.wq_b.output_quantizer                          TensorQuantizer(disabled)
+model.layers.30.self_attn.indexer.wq_b.weight_quantizer                          TensorQuantizer(disabled)
+model.layers.30.self_attn.indexer.wk.input_quantizer                             TensorQuantizer(disabled)
+model.layers.30.self_attn.indexer.wk.output_quantizer                            TensorQuantizer(disabled)
+model.layers.30.self_attn.indexer.wk.weight_quantizer                            TensorQuantizer(disabled)
+model.layers.30.self_attn.indexer.k_norm.input_quantizer                         TensorQuantizer(disabled)
+model.layers.30.self_attn.indexer.k_norm.output_quantizer                        TensorQuantizer(disabled)
+model.layers.30.self_attn.indexer.weights_proj.input_quantizer                   TensorQuantizer(disabled)
+model.layers.30.self_attn.indexer.weights_proj.output_quantizer                  TensorQuantizer(disabled)
+model.layers.30.self_attn.indexer.weights_proj.weight_quantizer                  TensorQuantizer(disabled)
+model.layers.30.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.30.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.30.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.30.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.30.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e+00 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.33e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.20e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.36e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.75e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.21e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.68e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.29e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.60e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.83e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.91e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.83e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.23e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.87e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.50e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.34e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.25e+01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.13e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.40e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.35e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.54e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.98e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.01e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.06e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.20e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.54e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.97e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.11e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.03e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.97e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.73e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.30e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.31e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.39e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.96e-01 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.40e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.30.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.30.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.30.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.30.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.30.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.30.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.30.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.30.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.30.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.31.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.31.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.31.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.31.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.31.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.31.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.31.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.31.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.31.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.31.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.31.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.31.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.31.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.31.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.31.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.31.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.31.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.31.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.31.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.31.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e+00 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.92e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.47e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.03e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.58e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.28e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.99e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.59e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.62e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.41e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.42e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.86e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.91e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.18e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.57e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.99e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.77e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.38e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.10e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e+01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.91e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.35e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.40e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.26e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.40e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.81e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.25e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.30e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.87e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.78e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.41e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.34e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.23e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.40e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.40e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.31.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.31.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.31.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.31.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.31.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.31.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.31.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.31.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.31.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.32.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.32.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.32.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.32.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.32.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.32.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.32.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.32.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.32.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.32.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.32.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.32.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.32.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.32.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.32.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.32.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.32.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.32.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.32.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.32.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e+00 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.87e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.50e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.23e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.99e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.12e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.89e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.67e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.77e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.05e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.83e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.05e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.26e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.04e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.05e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.21e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.35e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.34e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.25e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.52e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.73e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.22e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.28e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.09e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.75e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.56e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.19e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.99e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.04e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.46e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.73e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.87e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.79e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.70e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.18e+01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.93e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.81e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.94e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.86e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.40e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.40e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.40e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.75e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.04e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.30e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.97e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.03e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.31e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.92e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.02e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.00e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.24e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.10e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.16e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.02e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.89e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.70e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.11e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.78e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.36e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.06e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.32.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.32.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.32.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.32.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.32.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.32.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.32.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.32.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.32.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.33.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.33.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.33.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.33.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.33.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.33.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.33.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.33.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.33.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.33.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.33.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.33.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.33.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.33.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.33.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.33.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.33.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.33.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.33.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.33.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e+00 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.38e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.93e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.02e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.08e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.83e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.47e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.35e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.95e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.60e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.49e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.16e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.34e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.54e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.04e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.90e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.07e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.96e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.99e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.64e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.90e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.79e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.86e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.75e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.62e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.71e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.24e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.09e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.97e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.14e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.93e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.15e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e+01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.63e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.35e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.35e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.40e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.40e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.62e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.14e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.91e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.33.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.33.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.33.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.33.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.33.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.33.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.33.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.33.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.33.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.34.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.34.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.34.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.34.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.34.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.34.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.34.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.34.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.34.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.34.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.34.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.34.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.34.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.34.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.34.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.34.self_attn.indexer.wq_b.input_quantizer                           TensorQuantizer(disabled)
+model.layers.34.self_attn.indexer.wq_b.output_quantizer                          TensorQuantizer(disabled)
+model.layers.34.self_attn.indexer.wq_b.weight_quantizer                          TensorQuantizer(disabled)
+model.layers.34.self_attn.indexer.wk.input_quantizer                             TensorQuantizer(disabled)
+model.layers.34.self_attn.indexer.wk.output_quantizer                            TensorQuantizer(disabled)
+model.layers.34.self_attn.indexer.wk.weight_quantizer                            TensorQuantizer(disabled)
+model.layers.34.self_attn.indexer.k_norm.input_quantizer                         TensorQuantizer(disabled)
+model.layers.34.self_attn.indexer.k_norm.output_quantizer                        TensorQuantizer(disabled)
+model.layers.34.self_attn.indexer.weights_proj.input_quantizer                   TensorQuantizer(disabled)
+model.layers.34.self_attn.indexer.weights_proj.output_quantizer                  TensorQuantizer(disabled)
+model.layers.34.self_attn.indexer.weights_proj.weight_quantizer                  TensorQuantizer(disabled)
+model.layers.34.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.34.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.34.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.34.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.34.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e+00 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.97e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.84e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.97e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.81e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.27e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.78e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.48e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.99e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.05e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.71e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.81e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.22e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.06e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.20e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.11e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.32e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.07e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.25e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.83e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.59e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.89e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.32e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.14e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.38e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.29e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.12e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.07e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.91e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.86e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.14e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.91e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.38e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.10e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.87e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.27e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.76e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.89e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.45e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.05e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.12e+01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.26e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.28e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.02e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.05e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.40e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.35e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.25e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.38e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.80e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.57e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.16e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.73e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.68e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.84e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.93e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.40e-02 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.34.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.34.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.34.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.34.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.34.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.34.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.34.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.34.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.34.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.35.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.35.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.35.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.35.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.35.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.35.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.35.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.35.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.35.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.35.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.35.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.35.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.35.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.35.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.35.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.35.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.35.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.35.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.35.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.35.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e+00 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.65e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.88e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.50e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.85e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.26e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.85e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.15e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.81e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.05e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.20e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.12e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.92e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.26e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.89e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.20e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.12e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.92e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.54e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.49e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.97e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.12e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.23e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.38e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.52e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.23e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.78e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.16e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.73e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.07e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.67e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.05e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.67e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e+01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.32e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.79e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.03e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.85e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.01e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.30e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.34e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.35e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.39e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.29e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.05e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.92e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.40e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.43e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.40e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.84e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.40e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.44e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.13e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.30e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.37e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.91e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.01e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.97e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.23e-01 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.35.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.35.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.35.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.35.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.35.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.35.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.35.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.35.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.35.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.36.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.36.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.36.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.36.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.36.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.36.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.36.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.36.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.36.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.36.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.36.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.36.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.36.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.36.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.36.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.36.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.36.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.36.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.36.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.36.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.79e+00 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.77e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.73e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.90e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.79e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.52e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.79e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.34e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.94e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.25e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.00e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.42e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.16e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.58e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.44e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.70e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.47e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.20e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.20e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.18e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.83e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.73e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.16e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.55e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.78e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.80e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.59e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.48e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.77e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.95e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.94e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.73e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.59e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.91e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.32e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.94e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.37e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.03e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.29e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.20e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.24e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.01e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.95e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e+01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.95e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.94e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.07e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.14e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.40e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.84e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.92e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.77e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.44e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.03e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.00e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.04e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.29e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.77e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.35e-02 calibrator=MaxCalibrator quant)
+model.layers.36.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.36.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.36.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.36.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.36.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.36.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.36.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.36.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.36.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.37.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.37.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.37.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.37.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.37.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.37.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.37.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.37.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.37.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.37.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.37.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.37.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.37.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.37.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.37.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.37.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.37.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.37.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.37.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.37.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.91e+00 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.86e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.47e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.58e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.56e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.10e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.43e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.71e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.65e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.79e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.80e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.01e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.76e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.79e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.45e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.77e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.95e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.99e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.77e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.30e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.25e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.79e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.03e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.14e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.10e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.46e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.79e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.07e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.62e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.98e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.36e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.26e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.96e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.26e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.95e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.54e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.86e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.05e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.37e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e+01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.34e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.30e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.83e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.75e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.18e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.34e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.45e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.40e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.19e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.37.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.37.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.37.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.37.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.37.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.37.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.37.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.37.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.37.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.38.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.38.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.38.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.38.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.38.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.38.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.38.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.38.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.38.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.38.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.38.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.38.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.38.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.38.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.38.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.38.self_attn.indexer.wq_b.input_quantizer                           TensorQuantizer(disabled)
+model.layers.38.self_attn.indexer.wq_b.output_quantizer                          TensorQuantizer(disabled)
+model.layers.38.self_attn.indexer.wq_b.weight_quantizer                          TensorQuantizer(disabled)
+model.layers.38.self_attn.indexer.wk.input_quantizer                             TensorQuantizer(disabled)
+model.layers.38.self_attn.indexer.wk.output_quantizer                            TensorQuantizer(disabled)
+model.layers.38.self_attn.indexer.wk.weight_quantizer                            TensorQuantizer(disabled)
+model.layers.38.self_attn.indexer.k_norm.input_quantizer                         TensorQuantizer(disabled)
+model.layers.38.self_attn.indexer.k_norm.output_quantizer                        TensorQuantizer(disabled)
+model.layers.38.self_attn.indexer.weights_proj.input_quantizer                   TensorQuantizer(disabled)
+model.layers.38.self_attn.indexer.weights_proj.output_quantizer                  TensorQuantizer(disabled)
+model.layers.38.self_attn.indexer.weights_proj.weight_quantizer                  TensorQuantizer(disabled)
+model.layers.38.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.38.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.38.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.38.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.38.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.98e+00 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.20e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.05e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.31e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.00e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.50e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.86e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.21e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.20e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.45e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.27e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.07e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.66e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.85e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.91e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.29e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.52e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.06e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.44e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.18e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.97e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.21e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.32e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.46e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.99e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.44e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.82e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.58e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.32e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.75e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.22e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.42e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.01e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.71e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.07e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.99e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.20e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.12e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.14e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.99e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.84e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.03e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.40e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.12e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.03e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.60e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.86e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.98e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.02e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.02e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.61e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.89e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.07e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.95e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.75e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.12e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.71e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.22e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.81e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e+01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.06e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.97e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.36e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.98e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.86e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.76e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.79e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.78e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.91e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.99e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.91e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.77e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.08e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.54e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.38.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.38.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.38.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.38.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.38.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.38.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.38.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.38.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.38.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.39.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.39.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.39.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.39.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.39.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.39.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.39.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.39.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.39.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.39.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.39.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.39.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.39.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.39.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.39.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.39.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.39.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.39.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.39.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.39.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.12e+00 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.79e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.77e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.62e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.78e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.69e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.79e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.24e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.51e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.03e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.29e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.20e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.12e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.27e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.75e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.41e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.06e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.46e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.95e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.25e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.81e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.38e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.54e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.23e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.09e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.87e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.87e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.99e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.95e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.89e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.76e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.26e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.22e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.03e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.34e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.77e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.82e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.62e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.42e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.52e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.89e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.46e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.02e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.01e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.52e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.48e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.48e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.66e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.61e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.73e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.82e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.35e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e+01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.01e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.20e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.05e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.36e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.93e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.94e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.13e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.82e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.78e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.75e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.77e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.92e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.35e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.11e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.40e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.60e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.39.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.39.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.39.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.39.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.39.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.39.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.39.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.39.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.39.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.40.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.40.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.40.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.40.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.40.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.40.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.40.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.40.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.40.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.40.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.40.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.40.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.40.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.40.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.40.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.40.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.40.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.40.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.40.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.40.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.25e+00 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.84e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.15e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.22e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.68e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.99e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.93e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.24e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.06e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.03e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.33e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.40e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.28e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.95e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.25e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.93e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.97e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.04e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.56e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.14e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.36e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.81e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.93e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.05e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.26e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.79e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.28e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.12e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.47e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.34e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.27e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.91e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.71e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.59e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.27e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.75e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.13e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.70e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.15e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.71e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.95e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.46e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.79e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.29e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.93e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.18e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.21e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.97e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.22e+01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.98e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.91e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.41e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.40e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.97e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.29e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.23e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.78e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.76e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.32e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.98e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.25e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.40.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.40.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.40.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.40.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.40.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.40.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.40.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.40.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.40.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.41.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.41.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.41.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.41.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.41.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.41.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.41.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.41.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.41.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.41.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.41.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.41.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.41.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.41.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.41.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.41.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.41.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.41.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.41.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.41.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.41e+00 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.77e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.87e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.75e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.79e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.20e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.01e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.09e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.94e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.77e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.79e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.28e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.41e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.53e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.00e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.43e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.08e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.27e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.09e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.30e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.32e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.92e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.62e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.87e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.58e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.56e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.79e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.99e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.92e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.75e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.86e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.60e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.33e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.13e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.30e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.66e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.77e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.42e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.41e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.26e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.93e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.11e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.95e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.16e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.77e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.54e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.16e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e+01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.30e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.98e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.05e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.85e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.50e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.35e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.97e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.86e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.77e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.90e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.38e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.40e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.94e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.80e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.35e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.32e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.84e-01 calibrator=MaxCalibrator quant)
+model.layers.41.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.41.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.41.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.41.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.41.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.41.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.41.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.41.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.41.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.42.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.42.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.42.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.42.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.42.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.42.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.42.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.42.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.42.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.42.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.42.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.42.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.42.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.42.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.42.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.42.self_attn.indexer.wq_b.input_quantizer                           TensorQuantizer(disabled)
+model.layers.42.self_attn.indexer.wq_b.output_quantizer                          TensorQuantizer(disabled)
+model.layers.42.self_attn.indexer.wq_b.weight_quantizer                          TensorQuantizer(disabled)
+model.layers.42.self_attn.indexer.wk.input_quantizer                             TensorQuantizer(disabled)
+model.layers.42.self_attn.indexer.wk.output_quantizer                            TensorQuantizer(disabled)
+model.layers.42.self_attn.indexer.wk.weight_quantizer                            TensorQuantizer(disabled)
+model.layers.42.self_attn.indexer.k_norm.input_quantizer                         TensorQuantizer(disabled)
+model.layers.42.self_attn.indexer.k_norm.output_quantizer                        TensorQuantizer(disabled)
+model.layers.42.self_attn.indexer.weights_proj.input_quantizer                   TensorQuantizer(disabled)
+model.layers.42.self_attn.indexer.weights_proj.output_quantizer                  TensorQuantizer(disabled)
+model.layers.42.self_attn.indexer.weights_proj.weight_quantizer                  TensorQuantizer(disabled)
+model.layers.42.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.42.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.42.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.42.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.42.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.58e+00 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.76e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.81e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.77e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.73e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.19e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.30e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.23e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.75e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.50e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.93e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.85e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.33e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.85e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.04e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.95e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.73e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.78e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.58e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.37e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.33e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.06e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.99e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.84e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.68e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.19e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.86e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.47e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.54e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.24e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.79e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.13e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.46e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.08e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.06e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.96e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.44e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.80e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.96e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.00e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.05e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.81e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.15e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.14e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.23e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.12e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.14e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.77e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.34e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.11e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e+01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.14e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.16e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.56e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.75e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.86e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.73e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.70e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.89e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.54e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.40e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.40e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.29e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.86e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.95e-01 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.42.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.42.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.42.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.42.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.42.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.42.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.42.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.42.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.42.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.43.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.43.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.43.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.43.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.43.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.43.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.43.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.43.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.43.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.43.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.43.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.43.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.43.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.43.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.43.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.43.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.43.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.43.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.43.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.43.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.42e+00 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.43e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.31e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.34e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.21e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.30e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.82e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.47e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.36e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.27e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.32e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.82e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.96e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.90e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.36e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.29e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.69e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.73e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.87e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.64e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.12e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.23e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.91e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.46e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.64e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.48e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.32e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.35e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.22e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.23e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.80e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.02e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.44e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.99e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.46e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.23e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.03e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.46e+01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.99e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.46e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.89e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.11e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.89e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.99e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.58e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.17e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.02e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.40e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.91e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.73e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.29e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.87e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.91e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.43.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.43.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.43.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.43.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.43.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.43.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.43.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.43.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.43.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.44.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.44.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.44.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.44.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.44.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.44.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.44.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.44.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.44.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.44.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.44.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.44.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.44.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.44.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.44.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.44.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.44.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.44.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.44.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.44.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.45e+00 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.82e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.26e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.20e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.55e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.83e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.76e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.24e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.93e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.36e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.52e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.99e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.40e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.02e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.96e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.90e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.53e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.07e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.96e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.35e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.97e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.91e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.08e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.34e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.22e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.34e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.00e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.85e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.56e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.98e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.17e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.28e+01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.24e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.16e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.73e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.62e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.73e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.11e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.04e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.12e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.83e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.87e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.87e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.93e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.64e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.82e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.03e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.07e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.54e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.42e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.03e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.18e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.26e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.40e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.40e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.36e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.20e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.04e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.08e-01 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.44.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.44.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.44.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.44.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.44.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.44.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.44.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.44.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.44.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.45.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.45.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.45.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.45.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.45.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.45.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.45.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.45.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.45.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.45.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.45.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.45.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.45.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.45.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.45.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.45.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.45.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.45.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.45.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.45.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.66e+00 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.82e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.77e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.83e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.24e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.23e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.38e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.33e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.48e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.85e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.34e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.60e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.23e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.11e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.85e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.24e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.30e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.52e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.79e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.04e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.08e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.60e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.87e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.23e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.30e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.45e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.29e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.24e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.90e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.79e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.16e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.79e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.32e+01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.37e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.23e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.39e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.83e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.20e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.02e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.25e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.01e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.16e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.90e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.49e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.30e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.95e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.28e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.09e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.37e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.40e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.34e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.77e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.81e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.08e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.45.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.45.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.45.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.45.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.45.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.45.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.45.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.45.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.45.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.46.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.46.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.46.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.46.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.46.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.46.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.46.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.46.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.46.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.46.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.46.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.46.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.46.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.46.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.46.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.46.self_attn.indexer.wq_b.input_quantizer                           TensorQuantizer(disabled)
+model.layers.46.self_attn.indexer.wq_b.output_quantizer                          TensorQuantizer(disabled)
+model.layers.46.self_attn.indexer.wq_b.weight_quantizer                          TensorQuantizer(disabled)
+model.layers.46.self_attn.indexer.wk.input_quantizer                             TensorQuantizer(disabled)
+model.layers.46.self_attn.indexer.wk.output_quantizer                            TensorQuantizer(disabled)
+model.layers.46.self_attn.indexer.wk.weight_quantizer                            TensorQuantizer(disabled)
+model.layers.46.self_attn.indexer.k_norm.input_quantizer                         TensorQuantizer(disabled)
+model.layers.46.self_attn.indexer.k_norm.output_quantizer                        TensorQuantizer(disabled)
+model.layers.46.self_attn.indexer.weights_proj.input_quantizer                   TensorQuantizer(disabled)
+model.layers.46.self_attn.indexer.weights_proj.output_quantizer                  TensorQuantizer(disabled)
+model.layers.46.self_attn.indexer.weights_proj.weight_quantizer                  TensorQuantizer(disabled)
+model.layers.46.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.46.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.46.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.46.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.46.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.77e+00 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.73e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.05e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.09e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.84e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.79e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.23e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.86e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.21e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.21e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.18e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.09e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.36e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.41e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.75e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.61e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.09e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.35e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.10e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.70e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.80e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.03e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.54e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.13e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.95e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.16e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.46e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.13e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.28e+01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.24e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.18e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.46e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.50e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.97e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.45e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.79e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.43e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.52e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.76e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.83e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.19e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.16e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.93e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.16e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.92e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.14e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.24e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.29e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.46.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.46.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.46.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.46.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.46.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.46.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.46.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.46.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.46.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.47.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.47.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.47.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.47.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.47.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.47.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.47.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.47.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.47.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.47.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.47.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.47.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.47.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.47.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.47.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.47.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.47.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.47.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.47.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.47.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.83e+00 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.29e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.75e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.50e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.35e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.24e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.06e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.95e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.48e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.80e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.79e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.01e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.28e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.16e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.77e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.96e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.52e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.94e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.90e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.26e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.81e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.99e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.02e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.90e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.83e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.82e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.22e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.06e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.44e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.45e+01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.35e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.30e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.78e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.30e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.92e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.20e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.12e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.01e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.90e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.02e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.40e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.41e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.14e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.28e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.87e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.81e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.29e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.26e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.09e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.05e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.07e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.64e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.30e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.19e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.11e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.80e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.89e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.14e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.42e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.10e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.20e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.99e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.47.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.47.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.47.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.47.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.47.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.47.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.47.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.47.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.47.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.48.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.48.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.48.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.48.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.48.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.48.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.48.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.48.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.48.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.48.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.48.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.48.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.48.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.48.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.48.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.48.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.48.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.48.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.48.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.48.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.73e+00 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.76e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.01e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.68e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.42e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.91e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.11e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.93e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.37e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.93e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.32e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.95e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.49e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.84e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.09e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.40e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.93e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.20e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.97e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.97e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.98e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.52e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.85e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.75e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.14e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.75e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.00e+01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.95e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.05e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.12e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.90e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.47e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.78e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.52e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.42e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.84e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.40e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.17e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.18e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.38e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.54e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.89e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.02e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.00e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.70e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.48.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.48.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.48.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.48.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.48.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.48.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.48.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.48.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.48.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.49.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.49.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.49.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.49.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.49.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.49.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.49.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.49.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.49.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.49.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.49.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.49.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.49.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.49.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.49.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.49.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.49.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.49.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.49.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.49.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.84e+00 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.20e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.46e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.08e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.97e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.12e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.95e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.85e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.86e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.47e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.85e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.83e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.58e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.40e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.07e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.09e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.29e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.66e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.84e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.01e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.84e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.50e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.52e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.09e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.10e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.28e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.32e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.49e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.34e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.42e+01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.11e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.82e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.07e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.94e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.91e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.93e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.31e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.71e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.60e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.75e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.80e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.30e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.83e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.91e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.17e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.98e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.78e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.79e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.29e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.61e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.82e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.92e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.92e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.00e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.75e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.92e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.49.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.49.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.49.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.49.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.49.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.49.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.49.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.49.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.49.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.50.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.50.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.50.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.50.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.50.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.50.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.50.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.50.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.50.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.50.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.50.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.50.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.50.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.50.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.50.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.50.self_attn.indexer.wq_b.input_quantizer                           TensorQuantizer(disabled)
+model.layers.50.self_attn.indexer.wq_b.output_quantizer                          TensorQuantizer(disabled)
+model.layers.50.self_attn.indexer.wq_b.weight_quantizer                          TensorQuantizer(disabled)
+model.layers.50.self_attn.indexer.wk.input_quantizer                             TensorQuantizer(disabled)
+model.layers.50.self_attn.indexer.wk.output_quantizer                            TensorQuantizer(disabled)
+model.layers.50.self_attn.indexer.wk.weight_quantizer                            TensorQuantizer(disabled)
+model.layers.50.self_attn.indexer.k_norm.input_quantizer                         TensorQuantizer(disabled)
+model.layers.50.self_attn.indexer.k_norm.output_quantizer                        TensorQuantizer(disabled)
+model.layers.50.self_attn.indexer.weights_proj.input_quantizer                   TensorQuantizer(disabled)
+model.layers.50.self_attn.indexer.weights_proj.output_quantizer                  TensorQuantizer(disabled)
+model.layers.50.self_attn.indexer.weights_proj.weight_quantizer                  TensorQuantizer(disabled)
+model.layers.50.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.50.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.50.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.50.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.50.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.20e+00 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.29e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.22e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.99e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.75e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.71e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.01e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.06e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.14e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.05e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.62e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.90e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.85e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.64e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.02e+01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.83e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.30e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.85e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.86e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.40e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.30e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.15e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.40e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.99e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.71e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.40e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.96e-01 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.50.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.50.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.50.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.50.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.50.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.50.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.50.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.50.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.50.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.51.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.51.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.51.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.51.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.51.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.51.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.51.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.51.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.51.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.51.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.51.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.51.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.51.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.51.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.51.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.51.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.51.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.51.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.51.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.51.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.06e+00 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.77e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.87e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.77e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.13e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.78e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.41e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.80e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.96e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.87e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.43e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.09e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.78e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.75e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.32e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.88e+01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.85e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.03e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.54e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.32e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.40e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.92e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.24e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.42e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.44e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.06e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.87e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.07e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.58e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.83e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.04e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.77e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.51.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.51.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.51.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.51.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.51.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.51.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.51.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.51.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.51.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.52.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.52.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.52.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.52.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.52.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.52.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.52.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.52.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.52.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.52.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.52.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.52.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.52.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.52.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.52.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.52.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.52.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.52.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.52.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.52.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.34e+00 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.77e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.94e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.83e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.10e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.95e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.79e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.99e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.86e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.08e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.80e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.84e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.13e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.09e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.70e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.82e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.22e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.30e+01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.01e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.73e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.80e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.10e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.40e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.96e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.26e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.33e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.91e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.91e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.31e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.30e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.52.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.52.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.52.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.52.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.52.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.52.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.52.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.52.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.52.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.53.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.53.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.53.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.53.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.53.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.53.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.53.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.53.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.53.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.53.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.53.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.53.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.53.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.53.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.53.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.53.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.53.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.53.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.53.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.53.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.06e+00 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.62e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.77e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.05e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.65e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.68e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.42e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.06e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.60e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.69e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.71e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.62e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.83e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.89e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.50e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.65e+01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.89e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.73e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.40e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.79e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.40e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.53.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.53.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.53.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.53.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.53.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.53.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.53.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.53.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.53.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.54.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.54.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.54.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.54.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.54.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.54.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.54.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.54.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.54.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.54.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.54.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.54.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.54.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.54.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.54.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.54.self_attn.indexer.wq_b.input_quantizer                           TensorQuantizer(disabled)
+model.layers.54.self_attn.indexer.wq_b.output_quantizer                          TensorQuantizer(disabled)
+model.layers.54.self_attn.indexer.wq_b.weight_quantizer                          TensorQuantizer(disabled)
+model.layers.54.self_attn.indexer.wk.input_quantizer                             TensorQuantizer(disabled)
+model.layers.54.self_attn.indexer.wk.output_quantizer                            TensorQuantizer(disabled)
+model.layers.54.self_attn.indexer.wk.weight_quantizer                            TensorQuantizer(disabled)
+model.layers.54.self_attn.indexer.k_norm.input_quantizer                         TensorQuantizer(disabled)
+model.layers.54.self_attn.indexer.k_norm.output_quantizer                        TensorQuantizer(disabled)
+model.layers.54.self_attn.indexer.weights_proj.input_quantizer                   TensorQuantizer(disabled)
+model.layers.54.self_attn.indexer.weights_proj.output_quantizer                  TensorQuantizer(disabled)
+model.layers.54.self_attn.indexer.weights_proj.weight_quantizer                  TensorQuantizer(disabled)
+model.layers.54.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.54.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.54.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.54.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.54.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.94e+00 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.40e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.01e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.85e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.33e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.04e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.64e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.75e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.82e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.92e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.16e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.33e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.39e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.83e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.07e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.75e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.04e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.12e+01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.07e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.79e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.04e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.08e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.36e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.58e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.56e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.40e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.19e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.59e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.89e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.05e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.79e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.54.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.54.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.54.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.54.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.54.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.54.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.54.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.54.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.54.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.55.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.55.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.55.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.55.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.55.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.55.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.55.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.55.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.55.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.55.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.55.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.55.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.55.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.55.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.55.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.55.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.55.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.55.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.55.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.55.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.95e+00 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.77e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.86e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.01e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.32e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.85e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.88e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.22e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.14e+01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.01e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.30e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.59e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.76e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.03e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.78e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.96e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.83e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.30e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.34e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.07e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.55.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.55.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.55.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.55.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.55.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.55.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.55.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.55.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.55.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.56.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.56.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.56.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.56.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.56.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.56.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.56.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.56.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.56.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.56.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.56.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.56.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.56.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.56.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.56.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.56.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.56.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.56.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.56.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.56.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.12e+00 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.78e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.78e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.36e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.75e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.08e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.86e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.87e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.70e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.85e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.58e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.03e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.12e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.25e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.08e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.86e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.31e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.65e+01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.06e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.79e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.30e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.90e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.80e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.49e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.56.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.56.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.56.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.56.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.56.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.56.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.56.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.56.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.56.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.57.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.57.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.57.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.57.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.57.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.57.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.57.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.57.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.57.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.57.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.57.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.57.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.57.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.57.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.57.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.57.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.57.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.57.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.57.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.57.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.22e+00 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.11e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.36e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.25e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.24e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.26e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.15e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.46e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.36e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.95e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.11e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.47e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.80e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.30e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.96e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.09e+01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.40e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.45e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.25e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.82e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.38e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.31e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.13e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.57.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.57.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.57.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.57.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.57.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.57.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.57.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.57.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.57.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.58.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.58.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.58.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.58.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.58.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.58.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.58.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.58.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.58.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.58.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.58.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.58.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.58.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.58.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.58.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.58.self_attn.indexer.wq_b.input_quantizer                           TensorQuantizer(disabled)
+model.layers.58.self_attn.indexer.wq_b.output_quantizer                          TensorQuantizer(disabled)
+model.layers.58.self_attn.indexer.wq_b.weight_quantizer                          TensorQuantizer(disabled)
+model.layers.58.self_attn.indexer.wk.input_quantizer                             TensorQuantizer(disabled)
+model.layers.58.self_attn.indexer.wk.output_quantizer                            TensorQuantizer(disabled)
+model.layers.58.self_attn.indexer.wk.weight_quantizer                            TensorQuantizer(disabled)
+model.layers.58.self_attn.indexer.k_norm.input_quantizer                         TensorQuantizer(disabled)
+model.layers.58.self_attn.indexer.k_norm.output_quantizer                        TensorQuantizer(disabled)
+model.layers.58.self_attn.indexer.weights_proj.input_quantizer                   TensorQuantizer(disabled)
+model.layers.58.self_attn.indexer.weights_proj.output_quantizer                  TensorQuantizer(disabled)
+model.layers.58.self_attn.indexer.weights_proj.weight_quantizer                  TensorQuantizer(disabled)
+model.layers.58.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.58.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.58.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.58.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.58.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.42e+00 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.42e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.08e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.20e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.26e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.17e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.99e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.16e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.89e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.96e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.00e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.84e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.21e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.40e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.05e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.78e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.79e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.99e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.11e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.08e+01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.82e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.32e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.91e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.78e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.91e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.93e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.18e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.06e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.29e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.04e-01 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.58.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.58.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.58.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.58.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.58.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.58.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.58.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.58.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.58.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.59.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.59.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.59.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.59.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.59.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.59.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.59.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.59.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.59.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.59.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.59.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.59.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.59.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.59.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.59.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.59.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.59.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.59.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.59.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.59.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.12e+00 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.73e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.12e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.41e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.78e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.13e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.87e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.05e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.79e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.52e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.14e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.56e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.91e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.84e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.38e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.03e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.64e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.03e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.93e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.22e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.95e+01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.79e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.75e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.48e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.58e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.27e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.12e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.03e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.30e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.35e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.85e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.30e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.78e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.76e-01 calibrator=MaxCalibrator quant)
+model.layers.59.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.59.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.59.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.59.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.59.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.59.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.59.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.59.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.59.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.60.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.60.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.60.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.60.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.60.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.60.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.60.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.60.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.60.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.60.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.60.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.60.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.60.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.60.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.60.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.60.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.60.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.60.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.60.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.60.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.25e+00 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.39e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.77e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.86e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.77e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.54e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.22e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.43e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.22e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.58e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.84e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.75e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.48e+01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.64e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.89e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.04e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.47e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.82e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.15e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.98e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.83e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.93e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.32e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.86e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.70e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.40e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.60.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.60.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.60.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.60.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.60.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.60.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.60.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.60.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.60.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.61.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.61.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.61.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.61.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.61.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.61.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.61.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.61.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.61.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.61.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.61.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.61.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.61.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.61.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.61.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.61.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.61.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.61.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.61.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.61.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.34e+00 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.95e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.01e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.98e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.97e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.96e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.83e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.56e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.98e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.94e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.78e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.31e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.35e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.94e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.82e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.47e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.71e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.18e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.00e+01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.14e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.31e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.99e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.73e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.30e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.42e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.37e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.66e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.15e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.17e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.73e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.22e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.67e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.61.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.61.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.61.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.61.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.61.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.61.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.61.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.61.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.61.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.62.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.62.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.62.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.62.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.62.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.62.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.62.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.62.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.62.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.62.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.62.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.62.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.62.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.62.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.62.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.62.self_attn.indexer.wq_b.input_quantizer                           TensorQuantizer(disabled)
+model.layers.62.self_attn.indexer.wq_b.output_quantizer                          TensorQuantizer(disabled)
+model.layers.62.self_attn.indexer.wq_b.weight_quantizer                          TensorQuantizer(disabled)
+model.layers.62.self_attn.indexer.wk.input_quantizer                             TensorQuantizer(disabled)
+model.layers.62.self_attn.indexer.wk.output_quantizer                            TensorQuantizer(disabled)
+model.layers.62.self_attn.indexer.wk.weight_quantizer                            TensorQuantizer(disabled)
+model.layers.62.self_attn.indexer.k_norm.input_quantizer                         TensorQuantizer(disabled)
+model.layers.62.self_attn.indexer.k_norm.output_quantizer                        TensorQuantizer(disabled)
+model.layers.62.self_attn.indexer.weights_proj.input_quantizer                   TensorQuantizer(disabled)
+model.layers.62.self_attn.indexer.weights_proj.output_quantizer                  TensorQuantizer(disabled)
+model.layers.62.self_attn.indexer.weights_proj.weight_quantizer                  TensorQuantizer(disabled)
+model.layers.62.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.62.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.62.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.62.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.62.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.42e+00 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.98e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.96e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.90e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.14e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.76e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.87e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.77e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.82e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.05e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.89e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.82e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.79e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.97e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.78e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.04e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.78e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.30e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.27e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.96e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.42e+01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.77e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.35e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.52e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.43e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.12e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.93e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.40e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.80e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.62e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.43e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.84e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.33e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.62.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.62.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.62.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.62.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.62.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.62.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.62.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.62.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.62.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.63.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.63.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.63.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.63.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.63.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.63.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.63.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.63.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.63.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.63.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.63.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.63.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.63.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.63.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.63.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.63.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.63.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.63.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.63.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.63.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.47e+00 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.55e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.58e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.97e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.65e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.93e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.95e+00 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.77e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.97e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.29e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.28e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.52e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.67e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.85e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.97e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.22e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.82e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.73e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.95e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.20e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.50e+01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.37e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.86e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.75e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.76e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.11e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.02e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.91e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.38e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.57e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.95e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.00e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.85e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.50e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.34e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.63.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.63.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.63.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.63.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.63.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.63.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.63.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.63.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.63.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.64.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.64.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.64.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.64.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.64.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.64.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.64.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.64.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.64.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.64.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.64.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.64.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.64.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.64.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.64.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.64.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.64.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.64.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.64.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.64.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.53e+00 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.95e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.22e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.43e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.57e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.64e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.90e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.56e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.83e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.73e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.89e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.10e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.77e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.73e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.57e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.90e+01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.09e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.68e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.83e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.82e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.27e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.11e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.14e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.12e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.46e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.95e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.90e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.40e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.49e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.06e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.64.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.64.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.64.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.64.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.64.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.64.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.64.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.64.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.64.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.65.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.65.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.65.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.65.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.65.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.65.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.65.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.65.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.65.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.65.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.65.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.65.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.65.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.65.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.65.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.65.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.65.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.65.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.65.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.65.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.09e+00 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.67e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.29e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.59e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.97e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.87e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.92e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.16e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.50e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.98e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.03e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.76e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.08e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.92e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.28e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.76e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.79e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.38e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.06e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.72e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.20e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.70e+01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.52e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.54e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.83e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.93e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.06e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.32e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.09e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.95e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.57e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.29e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.09e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.07e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.99e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.95e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.23e-01 calibrator=MaxCalibrator quant)
+model.layers.65.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.65.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.65.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.65.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.65.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.65.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.65.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.65.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.65.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.66.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.66.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.66.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.66.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.66.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.66.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.66.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.66.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.66.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.66.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.66.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.66.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.66.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.66.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.66.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.66.self_attn.indexer.wq_b.input_quantizer                           TensorQuantizer(disabled)
+model.layers.66.self_attn.indexer.wq_b.output_quantizer                          TensorQuantizer(disabled)
+model.layers.66.self_attn.indexer.wq_b.weight_quantizer                          TensorQuantizer(disabled)
+model.layers.66.self_attn.indexer.wk.input_quantizer                             TensorQuantizer(disabled)
+model.layers.66.self_attn.indexer.wk.output_quantizer                            TensorQuantizer(disabled)
+model.layers.66.self_attn.indexer.wk.weight_quantizer                            TensorQuantizer(disabled)
+model.layers.66.self_attn.indexer.k_norm.input_quantizer                         TensorQuantizer(disabled)
+model.layers.66.self_attn.indexer.k_norm.output_quantizer                        TensorQuantizer(disabled)
+model.layers.66.self_attn.indexer.weights_proj.input_quantizer                   TensorQuantizer(disabled)
+model.layers.66.self_attn.indexer.weights_proj.output_quantizer                  TensorQuantizer(disabled)
+model.layers.66.self_attn.indexer.weights_proj.weight_quantizer                  TensorQuantizer(disabled)
+model.layers.66.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.66.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.66.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.66.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.66.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.89e+00 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.91e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.62e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.12e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.09e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.49e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.89e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.36e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.46e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.21e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.94e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.94e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.51e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.00e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.59e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.11e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.20e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.91e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.87e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.08e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.48e+01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.32e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.83e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.89e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.87e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.44e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.75e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.42e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.50e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.66.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.66.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.66.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.66.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.66.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.66.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.66.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.66.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.66.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.67.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.67.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.67.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.67.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.67.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.67.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.67.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.67.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.67.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.67.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.67.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.67.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.67.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.67.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.67.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.67.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.67.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.67.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.67.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.67.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.03e+00 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.23e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.20e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.87e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.20e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.89e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.38e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.90e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.73e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.87e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.76e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.14e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.14e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.76e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.91e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.02e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.98e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.36e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.84e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.43e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.09e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.00e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.07e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.40e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.82e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.74e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.67e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.08e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.85e+01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.50e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.73e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.29e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.45e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.54e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.82e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.41e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.21e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.16e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.76e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.19e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.67.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.67.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.67.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.67.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.67.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.67.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.67.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.67.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.67.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.68.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.68.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.68.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.68.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.68.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.68.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.68.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.68.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.68.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.68.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.68.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.68.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.68.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.68.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.68.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.68.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.68.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.68.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.68.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.68.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.75e+00 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.24e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.91e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.21e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.24e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.29e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e+00 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.92e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.90e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.98e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.97e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.75e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e+00 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.86e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.75e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.85e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.05e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.64e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.75e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.67e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.51e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.16e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.45e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.49e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.57e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.80e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.75e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.80e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.12e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.29e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.75e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.55e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.82e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.08e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.33e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.79e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.94e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.00e+01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.89e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.79e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.32e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.39e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.73e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.66e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.91e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.96e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.50e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.77e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.29e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.41e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.48e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.91e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.66e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.87e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.93e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.16e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.58e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.28e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.79e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.33e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.73e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.02e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.60e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.97e-01 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.68.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.68.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.68.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.68.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.68.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.68.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.68.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.68.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.68.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.69.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.69.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.69.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.69.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.69.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.69.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.69.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.69.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.69.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.69.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.69.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.69.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.69.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.69.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.69.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.69.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.69.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.69.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.69.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.69.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.64e+00 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.58e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.02e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.89e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.27e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.75e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.79e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.96e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.68e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.82e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.52e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.42e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.80e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.83e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.16e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.23e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.06e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.71e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.78e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.19e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.56e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.38e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.97e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.99e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.77e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.00e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.43e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.61e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.80e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.95e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.73e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.06e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.60e+01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.12e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.77e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.77e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.93e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.52e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.33e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.59e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.03e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.74e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.71e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.09e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.20e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.28e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.36e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.20e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.18e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.17e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.27e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.04e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.98e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.69.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.69.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.69.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.69.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.69.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.69.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.69.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.69.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.69.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.70.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.70.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.70.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.70.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.70.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.70.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.70.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.70.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.70.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.70.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.70.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.70.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.70.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.70.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.70.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.70.self_attn.indexer.wq_b.input_quantizer                           TensorQuantizer(disabled)
+model.layers.70.self_attn.indexer.wq_b.output_quantizer                          TensorQuantizer(disabled)
+model.layers.70.self_attn.indexer.wq_b.weight_quantizer                          TensorQuantizer(disabled)
+model.layers.70.self_attn.indexer.wk.input_quantizer                             TensorQuantizer(disabled)
+model.layers.70.self_attn.indexer.wk.output_quantizer                            TensorQuantizer(disabled)
+model.layers.70.self_attn.indexer.wk.weight_quantizer                            TensorQuantizer(disabled)
+model.layers.70.self_attn.indexer.k_norm.input_quantizer                         TensorQuantizer(disabled)
+model.layers.70.self_attn.indexer.k_norm.output_quantizer                        TensorQuantizer(disabled)
+model.layers.70.self_attn.indexer.weights_proj.input_quantizer                   TensorQuantizer(disabled)
+model.layers.70.self_attn.indexer.weights_proj.output_quantizer                  TensorQuantizer(disabled)
+model.layers.70.self_attn.indexer.weights_proj.weight_quantizer                  TensorQuantizer(disabled)
+model.layers.70.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.70.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.70.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.70.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.70.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.58e+00 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.07e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.00e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.75e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.24e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.78e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.96e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.88e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.65e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.50e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.11e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.59e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.39e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.98e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.46e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.06e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.39e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.03e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.10e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.20e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.71e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.75e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.87e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.77e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.87e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.07e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.95e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.33e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e+00 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.91e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.77e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.16e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.36e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e+00 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.56e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.70e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.04e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e+02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.16e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.21e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.86e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.32e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.39e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.29e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.96e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.03e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.03e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.43e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.89e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.16e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.86e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.87e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.57e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.27e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.78e-01 calibrator=MaxCalibrator quant)
+model.layers.70.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.70.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.70.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.70.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.70.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.70.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.70.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.70.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.70.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.71.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.71.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.71.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.71.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.71.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.71.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.71.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.71.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.71.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.71.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.71.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.71.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.71.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.71.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.71.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.71.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.71.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.71.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.71.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.71.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.61e+00 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.09e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.01e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.89e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.25e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.81e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.95e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.27e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.86e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.30e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.78e+00 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.79e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.45e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.82e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.12e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.50e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.64e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.32e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.52e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.11e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.07e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.93e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.77e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.91e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.24e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.38e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.91e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.34e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.06e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.36e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.80e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.77e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.15e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.26e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.23e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.46e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e+02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.18e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.81e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.98e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.70e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.73e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.49e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.41e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.45e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.87e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.96e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.00e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.15e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.12e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.00e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.75e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.17e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.11e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.99e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.71.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.71.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.71.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.71.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.71.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.71.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.71.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.71.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.71.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.72.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.72.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.72.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.72.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.72.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.72.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.72.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.72.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.72.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.72.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.72.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.72.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.72.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.72.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.72.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.72.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.72.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.72.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.72.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.72.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.97e+00 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e+00 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.64e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.70e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.83e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.45e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.66e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.73e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.47e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.47e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.12e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.42e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.62e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.07e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.87e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.76e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.54e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.57e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.12e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.66e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.33e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.38e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.31e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.77e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.20e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.75e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.00e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e+02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.77e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.79e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.09e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.27e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.70e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.89e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.80e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.18e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.85e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.85e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.43e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.24e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.25e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.80e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.08e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.30e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.84e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.80e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.84e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.37e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.10e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.20e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.72.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.72.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.72.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.72.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.72.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.72.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.72.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.72.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.72.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.73.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.73.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.73.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.73.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.73.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.73.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.73.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.73.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.73.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.73.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.73.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.73.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.73.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.73.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.73.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.73.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.73.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.73.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.73.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.73.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.72e+00 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.65e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.09e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.31e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.26e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.03e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.94e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.79e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.24e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.20e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.14e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.79e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.90e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.89e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.75e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.43e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.24e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.79e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.23e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.35e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.13e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.98e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.83e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.47e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.09e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.93e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.17e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.10e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.06e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.38e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.09e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.39e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.81e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.95e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.79e+02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.10e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.89e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.85e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.04e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.86e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.26e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.64e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.03e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.17e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.85e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.02e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.08e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.20e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.63e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.92e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.71e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.23e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.75e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.99e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.27e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.46e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.46e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.75e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.66e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.91e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.29e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.13e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.73e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.77e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.37e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.75e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.82e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.79e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.85e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.73e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.12e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.79e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.00e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.94e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.73.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.73.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.73.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.73.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.73.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.73.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.73.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.73.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.73.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.74.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.74.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.74.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.74.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.74.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.74.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.74.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.74.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.74.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.74.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.74.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.74.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.74.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.74.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.74.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.74.self_attn.indexer.wq_b.input_quantizer                           TensorQuantizer(disabled)
+model.layers.74.self_attn.indexer.wq_b.output_quantizer                          TensorQuantizer(disabled)
+model.layers.74.self_attn.indexer.wq_b.weight_quantizer                          TensorQuantizer(disabled)
+model.layers.74.self_attn.indexer.wk.input_quantizer                             TensorQuantizer(disabled)
+model.layers.74.self_attn.indexer.wk.output_quantizer                            TensorQuantizer(disabled)
+model.layers.74.self_attn.indexer.wk.weight_quantizer                            TensorQuantizer(disabled)
+model.layers.74.self_attn.indexer.k_norm.input_quantizer                         TensorQuantizer(disabled)
+model.layers.74.self_attn.indexer.k_norm.output_quantizer                        TensorQuantizer(disabled)
+model.layers.74.self_attn.indexer.weights_proj.input_quantizer                   TensorQuantizer(disabled)
+model.layers.74.self_attn.indexer.weights_proj.output_quantizer                  TensorQuantizer(disabled)
+model.layers.74.self_attn.indexer.weights_proj.weight_quantizer                  TensorQuantizer(disabled)
+model.layers.74.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.74.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.74.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.74.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.74.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.78e+00 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.62e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.87e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.05e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.95e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.77e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.94e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.26e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.89e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.22e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.73e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.29e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.63e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.46e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.11e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.36e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.50e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.13e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.42e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.01e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.93e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.73e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.81e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.62e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.30e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.94e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.49e+02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.62e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.98e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.89e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.52e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.34e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.79e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.11e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.99e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.15e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.05e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.11e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.15e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.48e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.16e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.30e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.69e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.27e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.98e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.70e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.02e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.87e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.21e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.79e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.27e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.73e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.87e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.08e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.01e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.01e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.64e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.79e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.86e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.37e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.44e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.45e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.38e-01 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.74.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.74.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.74.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.74.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.74.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.74.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.74.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.74.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.74.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.75.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.75.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.75.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.75.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.75.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.75.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.75.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.75.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.75.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.75.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.75.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.75.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.75.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.75.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.75.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.75.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.75.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.75.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.75.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.75.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.91e+00 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.16e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.22e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.08e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.49e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.22e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.38e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.04e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.19e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.37e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.52e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.75e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.28e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.83e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.09e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.76e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.84e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.79e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.94e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.03e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.03e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.83e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.40e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.09e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.20e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.71e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.31e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.95e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.47e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.24e+02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.85e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.84e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.52e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.82e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.03e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.17e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.93e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.58e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.38e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.75e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.20e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.45e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.15e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.46e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.72e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.06e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.59e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.95e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.77e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.83e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.24e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.05e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.07e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.50e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.42e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.76e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.77e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.09e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.73e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.71e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.76e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.18e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.07e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.36e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.47e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.73e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.75e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.23e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.75.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.75.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.75.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.75.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.75.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.75.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.75.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.75.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.75.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.76.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.76.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.76.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.76.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.76.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.76.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.76.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.76.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.76.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.76.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.76.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.76.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.76.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.76.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.76.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.76.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.76.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.76.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.76.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.76.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.50e+00 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.19e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.76e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.99e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.96e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.92e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.14e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.78e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.70e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.66e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.52e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.94e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.73e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.16e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.11e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.92e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.93e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.44e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.98e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.62e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.80e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.11e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.39e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.34e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.09e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.14e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.73e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.09e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.76e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.77e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.88e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.80e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.08e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.78e+02 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.05e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.32e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.35e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.26e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.01e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.79e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.42e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.44e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.93e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.79e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.42e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.91e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.96e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.91e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.98e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.75e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.86e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.71e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.24e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.12e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.75e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.31e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.89e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.60e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.95e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.57e-02 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.01e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.84e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.00e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.79e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.97e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.27e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.23e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.18e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.06e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.96e-02 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.74e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.52e-02 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.44e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.57e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.77e-02 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.07e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.67e-02 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.00e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.03e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.33e-02 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.51e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.09e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.91e-02 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.34e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.62e-02 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.38e-02 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.86e-02 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.81e-02 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.07e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.76.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.76.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.76.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.76.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.76.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.76.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.76.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.76.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.76.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.77.self_attn.q_a_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.77.self_attn.q_a_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.77.self_attn.q_a_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.77.self_attn.q_b_proj.input_quantizer                               TensorQuantizer(disabled)
+model.layers.77.self_attn.q_b_proj.output_quantizer                              TensorQuantizer(disabled)
+model.layers.77.self_attn.q_b_proj.weight_quantizer                              TensorQuantizer(disabled)
+model.layers.77.self_attn.kv_a_proj_with_mqa.input_quantizer                     TensorQuantizer(disabled)
+model.layers.77.self_attn.kv_a_proj_with_mqa.output_quantizer                    TensorQuantizer(disabled)
+model.layers.77.self_attn.kv_a_proj_with_mqa.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.77.self_attn.kv_b_proj.input_quantizer                              TensorQuantizer(disabled)
+model.layers.77.self_attn.kv_b_proj.output_quantizer                             TensorQuantizer(disabled)
+model.layers.77.self_attn.kv_b_proj.weight_quantizer                             TensorQuantizer(disabled)
+model.layers.77.self_attn.o_proj.input_quantizer                                 TensorQuantizer(disabled)
+model.layers.77.self_attn.o_proj.output_quantizer                                TensorQuantizer(disabled)
+model.layers.77.self_attn.o_proj.weight_quantizer                                TensorQuantizer(disabled)
+model.layers.77.self_attn.q_bmm_quantizer                                        TensorQuantizer(disabled)
+model.layers.77.self_attn.k_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.77.self_attn.v_bmm_quantizer                                        TensorQuantizer((4, 3) bit fake per-tensor amax=dynamic calibrator=MaxCalibrator quant)
+model.layers.77.self_attn.softmax_quantizer                                      TensorQuantizer(disabled)
+model.layers.77.mlp.experts.gate_up_proj_input_quantizer                         TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.38e+00 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.0                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.1                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.2                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.3                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.16e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.4                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.5                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.6                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.7                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.8                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.9                     TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.10                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.11                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.12                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.05e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.13                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.17e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.14                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.15                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.16                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.17                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.18                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.75e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.19                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.20                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.21                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.89e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.22                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.12e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.23                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.24                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.82e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.25                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.26                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.89e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.27                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.33e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.28                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.86e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.29                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.30                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.31                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.32                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.35e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.33                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.92e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.34                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.33e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.35                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e+00 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.36                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.80e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.37                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.38                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.39                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.79e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.40                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.41                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.42                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.43                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.44                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e+00 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.45                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.46                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.80e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.47                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.01e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.48                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.20e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.49                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.89e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.50                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.51                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.52                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.53                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.73e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.54                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.55                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.63e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.56                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.14e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.57                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.58                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.59                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.60                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.22e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.61                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.30e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.62                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.63                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.64                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.80e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.65                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.52e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.66                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.32e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.67                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.91e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.68                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.58e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.69                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.70                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.71                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.72                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.73                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.74                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.54e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.75                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.76                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.77                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.78                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.79                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.80                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.81                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.82                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.83                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.84                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.85                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.86                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.87                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.88                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.89                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.90                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.30e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.91                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.92                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.93                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.94                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.95                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.96                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.78e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.97                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.98                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.79e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.99                    TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.100                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.68e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.101                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.102                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.103                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.15e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.104                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.105                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.56e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.106                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.41e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.107                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.108                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.109                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.86e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.110                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.111                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.112                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.19e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.113                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.12e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.114                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.75e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.115                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.116                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.75e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.117                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.118                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.85e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.119                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.120                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.121                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.122                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.09e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.123                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.45e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.124                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.125                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.95e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.126                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.72e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.127                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.128                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.129                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.130                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.131                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.132                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.22e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.133                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.93e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.134                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.15e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.135                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.36e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.136                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.137                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.138                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.139                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.36e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.140                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.141                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.142                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.56e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.143                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.144                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.145                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.82e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.146                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.147                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.148                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.149                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.150                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.151                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.152                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.153                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.154                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.155                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.156                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.157                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.86e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.158                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.159                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.160                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.20e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.161                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.162                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.163                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.95e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.164                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.165                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.26e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.166                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.89e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.167                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.52e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.168                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.44e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.169                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.170                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.48e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.171                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.172                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.173                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.174                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.175                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.176                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.177                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.82e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.178                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.179                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.180                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.181                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.182                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.183                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.51e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.184                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.185                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.82e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.186                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.12e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.187                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.188                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.49e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.189                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.47e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.190                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.191                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.22e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.192                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.193                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.17e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.194                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.195                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.39e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.196                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.197                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.42e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.198                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.199                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.200                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.00e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.201                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.21e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.202                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.203                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.204                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.205                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.11e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.206                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.207                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.20e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.208                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.80e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.209                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.50e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.210                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.60e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.211                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.45e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.212                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.66e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.213                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.11e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.214                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.215                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.216                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.217                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.97e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.218                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.88e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.219                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.10e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.220                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.31e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.221                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.84e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.222                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.12e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.223                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.224                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.40e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.225                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.226                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.227                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.30e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.228                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.19e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.229                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.230                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.231                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.22e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.232                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.77e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.233                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.234                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.43e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.235                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.236                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.237                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.238                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.239                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.240                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.97e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.241                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.242                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.01e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.243                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.64e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.244                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.44e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.245                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.246                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.73e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.247                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.93e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.248                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.90e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.249                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.250                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.251                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.252                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.253                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.254                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.95e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.gate_up_proj_weight_quantizers.255                   TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.99e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_input_quantizer                            TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.79e+03 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.0                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.52e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.1                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.17e+00 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.2                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.13e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.3                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.86e+00 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.4                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.51e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.5                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.37e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.6                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.70e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.7                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.30e+00 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.8                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.20e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.9                        TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.82e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.10                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.14e+00 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.11                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.02e+00 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.12                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.76e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.13                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.28e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.14                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.30e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.15                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.36e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.16                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.46e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.17                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.60e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.18                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.11e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.19                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.32e+00 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.20                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.29e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.21                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e+00 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.22                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.82e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.23                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.34e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.24                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.28e+00 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.25                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.67e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.26                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.75e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.27                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.01e+00 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.28                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.85e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.29                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.69e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.30                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.23e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.31                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.43e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.32                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.08e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.33                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.35e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.34                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.01e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.35                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.17e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.36                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.79e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.37                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.63e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.38                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.43e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.39                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.35e+00 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.40                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.41                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.71e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.42                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.46e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.43                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.05e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.44                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.79e+00 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.45                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.67e+00 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.46                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.12e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.47                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.20e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.48                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.51e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.49                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.71e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.50                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e+00 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.51                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.55e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.52                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.83e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.53                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.54                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.16e+00 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.55                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.14e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.56                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.91e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.57                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.22e+00 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.58                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.18e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.59                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.47e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.60                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.24e+00 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.61                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e+00 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.62                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.70e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.63                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.43e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.64                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.27e+00 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.65                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.52e+00 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.66                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.37e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.67                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.50e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.68                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.55e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.69                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.73e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.70                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.38e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.71                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.08e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.72                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.44e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.73                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.41e+00 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.74                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.38e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.75                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.55e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.76                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.28e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.77                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.78                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e+00 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.79                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.95e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.80                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.68e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.81                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.55e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.82                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.17e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.83                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.52e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.84                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.12e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.85                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.37e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.86                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.50e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.87                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.96e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.88                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.36e+00 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.89                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.70e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.90                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.15e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.91                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.00e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.92                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.76e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.93                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.33e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.94                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.49e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.95                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.98e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.96                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.82e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.97                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.51e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.98                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.16e+00 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.99                       TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.21e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.100                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.20e+00 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.101                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e+00 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.102                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.49e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.103                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.34e+00 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.104                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.54e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.105                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.93e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.106                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.16e+00 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.107                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.85e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.108                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.61e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.109                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.05e+00 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.110                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.93e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.111                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.11e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.112                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.25e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.113                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.86e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.114                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.49e+00 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.115                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.10e+00 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.116                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.98e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.117                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.89e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.118                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.40e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.119                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.04e+00 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.120                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.54e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.121                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.05e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.122                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.34e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.123                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.32e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.124                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.45e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.125                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.25e+00 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.126                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.03e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.127                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.93e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.128                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.71e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.129                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.97e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.130                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.60e+00 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.131                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.59e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.132                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.20e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.133                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.53e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.134                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e+00 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.135                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.01e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.136                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.73e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.137                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.02e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.138                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.10e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.139                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.36e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.140                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.95e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.141                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.61e+00 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.142                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.91e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.143                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.65e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.144                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.07e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.145                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.52e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.146                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.18e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.147                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.73e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.148                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.12e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.149                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.78e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.150                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.38e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.151                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.71e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.152                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.55e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.153                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.98e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.154                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.34e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.155                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.39e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.156                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e+00 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.157                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.22e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.158                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.05e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.159                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.32e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.160                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.77e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.161                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.84e+00 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.162                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.16e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.163                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.62e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.164                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.51e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.165                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.01e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.166                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.13e+00 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.167                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.30e+00 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.168                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e+00 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.169                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.33e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.170                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.44e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.171                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.79e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.172                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.35e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.173                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.93e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.174                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.14e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.175                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.29e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.176                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.51e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.177                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.93e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.178                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.12e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.179                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.43e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.180                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.93e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.181                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.32e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.182                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.71e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.183                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.52e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.184                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.91e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.185                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.15e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.186                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.41e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.187                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.23e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.188                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.28e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.189                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.00e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.190                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.75e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.191                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.82e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.192                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.20e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.193                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.18e+00 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.194                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.69e+00 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.195                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.89e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.196                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.71e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.197                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.55e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.198                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.66e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.199                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.85e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.200                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.47e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.201                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.66e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.202                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.52e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.203                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.81e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.204                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.20e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.205                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.64e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.206                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.11e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.207                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.44e+00 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.208                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.26e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.209                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.75e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.210                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.59e+00 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.211                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.07e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.212                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.72e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.213                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.59e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.214                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.22e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.215                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.12e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.216                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.08e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.217                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.71e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.218                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.12e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.219                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=7.81e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.220                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.73e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.221                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.65e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.222                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.55e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.223                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.95e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.224                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.13e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.225                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.58e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.226                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.08e+00 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.227                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.36e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.228                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.79e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.229                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.46e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.230                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=5.35e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.231                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.62e+00 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.232                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.53e+00 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.233                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.75e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.234                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.81e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.235                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=6.09e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.236                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.47e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.237                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.71e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.238                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=4.06e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.239                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.38e+00 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.240                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.03e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.241                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.55e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.242                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.23e+00 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.243                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.27e+00 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.244                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.14e+00 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.245                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.40e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.246                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=9.49e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.247                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=8.05e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.248                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.40e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.249                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.02e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.250                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.44e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.251                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.09e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.252                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=3.75e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.253                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.93e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.254                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=2.13e-01 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.experts.down_proj_weight_quantizers.255                      TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=1.09e+00 calibrator=MaxCalibrator quant)
+model.layers.77.mlp.shared_experts.gate_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.77.mlp.shared_experts.gate_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.77.mlp.shared_experts.gate_proj.weight_quantizer                    TensorQuantizer(disabled)
+model.layers.77.mlp.shared_experts.up_proj.input_quantizer                       TensorQuantizer(disabled)
+model.layers.77.mlp.shared_experts.up_proj.output_quantizer                      TensorQuantizer(disabled)
+model.layers.77.mlp.shared_experts.up_proj.weight_quantizer                      TensorQuantizer(disabled)
+model.layers.77.mlp.shared_experts.down_proj.input_quantizer                     TensorQuantizer(disabled)
+model.layers.77.mlp.shared_experts.down_proj.output_quantizer                    TensorQuantizer(disabled)
+model.layers.77.mlp.shared_experts.down_proj.weight_quantizer                    TensorQuantizer(disabled)
+lm_head.input_quantizer                                                          TensorQuantizer(disabled)
+lm_head.output_quantizer                                                         TensorQuantizer(disabled)
+lm_head.weight_quantizer                                                         TensorQuantizer(disabled)
+40971 TensorQuantizers found in model
diff --git a/README.md b/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..d445a1579e95a6e07707f17b95c6746cb83afac0
--- /dev/null
+++ b/README.md
@@ -0,0 +1,247 @@
+---
+pipeline_tag: text-generation
+base_model:
+- zai-org/GLM-5.2
+license: mit
+library_name: Model Optimizer
+tags:
+- nvidia
+- ModelOpt
+- GLM-5
+- quantized
+- 4-bit precision
+- FP4
+- fp4
+---
+
+# Model Overview
+
+## Description:
+The NVIDIA GLM-5.2 NVFP4 model is the quantized version of ZAI’s GLM-5.2 model, which is an auto-regressive language model that uses an optimized transformer architecture. GLM-5.2 is a Mixture-of-Experts (MoE) model for reasoning and coding that uses sparse attention (with an IndexShare indexer) to support a long context. For more information, please check [here](https://huggingface.co/zai-org/GLM-5.2). The NVIDIA GLM-5.2 NVFP4 model is quantized with [Model Optimizer](https://github.com/NVIDIA/Model-Optimizer).
+
+This model is ready for commercial or non-commercial use.  <br>
+
+### License/Terms of Use:
+**GOVERNING TERMS:** Use of the model is governed by the [MIT License](https://huggingface.co/datasets/choosealicense/licenses/blob/main/markdown/mit.md), same as the base model.
+
+### Deployment Geography:
+Global <br>
+
+### Use Case: <br>
+Developers looking to take off-the-shelf, pre-quantized models for deployment in AI Agent systems, chatbots, RAG systems, and other AI-powered applications. <br>
+
+### Release Date:  <br>
+Hugging Face 06/25/2026 via https://huggingface.co/nvidia/GLM-5.2-NVFP4 <br>
+
+## References
+Nvidia Model Optimizer: https://github.com/NVIDIA/Model-Optimizer
+
+## Model Architecture:
+**Architecture Type:** Transformers  <br>
+**Network Architecture:** GLM-5.2 (`GlmMoeDsaForCausalLM`) <br>
+**Number of Model Parameters:** 753B in total and 40B activated <br>
+
+## Input:
+**Input Type(s):** Text <br>
+**Input Format(s):** String <br>
+**Input Parameters:** One-Dimensional (1D) <br>
+**Other Properties Related to Input:** Context length up to 1M <br>
+
+## Output:
+**Output Type(s):** Text <br>
+**Output Format:** String <br>
+**Output Parameters:** 1D (One-Dimensional): Sequences <br>
+**Other Properties Related to Output:** None <br>
+
+Our AI models are designed and/or optimized to run on NVIDIA GPU-accelerated systems. By leveraging NVIDIA’s hardware (e.g. GPU cores) and software frameworks (e.g., CUDA libraries), the model achieves faster training and inference times compared to CPU-only solutions. <br>
+
+## Software Integration:
+**Supported Runtime Engine(s):** <br>
+* SGLang <br>
+* vLLM <br>
+
+**Supported Hardware Microarchitecture Compatibility:** <br>
+* NVIDIA Blackwell <br>
+
+**Preferred Operating System(s):** <br>
+* Linux <br>
+
+The integration of foundation and fine-tuned models into AI systems requires additional testing using use-case-specific data to ensure safe and effective deployment. Following the V-model methodology, iterative testing and validation at both unit and system levels are essential to mitigate risks, meet technical and functional requirements, and ensure compliance with safety and ethical standards before deployment.
+
+## Model Version(s):
+The model version is NVFP4 1.0 version and is quantized with nvidia-modelopt **v0.46.0**  <br>
+
+## Training, Testing, and Evaluation Datasets:
+We calibrated the model using the dataset noted below, and performed evaluation using the benchmarks noted under Evaluation Datasets.
+We did not perform training or testing for this Model Optimizer release. The methods noted under Training and Testing Datasets below represent the data collection and labeling methods used by the third-party to train and test the underlying model.<br>
+
+## Training Dataset:
+**Data Modality:** Undisclosed <br>
+**Data Collection Method by dataset:** Undisclosed <br>
+**Labeling Method by dataset:** Undisclosed<br>
+**Properties:** Undisclosed
+
+## Testing Dataset:
+**Data Collection Method by dataset:** Undisclosed <br>
+**Labeling Method by dataset:** Undisclosed <br>
+**Properties:** Undisclosed <br>
+
+## Evaluation Dataset:
+**Datasets:** GPQA Diamond, SciCode, IFBench, AA-LCR, τ²-Bench Telecom <br>
+**Data Collection Method by dataset:** Hybrid: Automated, Human <br>
+**Labeling Method by dataset:** Hybrid: Human, Automated <br>
+**Properties:** We evaluated the model on text-based reasoning, coding, long-context recall, and agentic tool-use benchmarks: GPQA Diamond contains 448 graduate-level multiple-choice questions written by domain experts in biology, physics, and chemistry; SciCode evaluates scientific coding capabilities; IFBench is a benchmark for evaluating instruction-following capabilities across diverse and structured task constraints; AA-LCR (Artificial Analysis Long Context Recall) evaluates a model's ability to accurately retrieve and recall information from long input contexts; τ²-Bench Telecom evaluates agentic tool-use and policy-adherence capabilities in dual-control telecom customer-service scenarios where the model interacts with a simulated user and external tools to resolve account issues. <br>
+
+## Inference:
+**Acceleration Engine:** SGLang, vLLM <br>
+**Test Hardware:** NVIDIA B200 <br>
+                   NVIDIA B300 <br>
+
+## Post Training Quantization
+This model was obtained by quantizing the weights and activations of GLM-5.2 to NVFP4 data type, ready for inference with SGLang and vLLM. Only the weights and activations of the linear operators within transformer blocks in MoE experts are quantized. The shared expert is not quantized.
+
+## Usage
+
+### SGLang
+
+This checkpoint was served with the latest SGLang image (`lmsysorg/sglang:latest`). GLM-5.2's `glm_moe_dsa` architecture requires `transformers>=5.3.0`, which we installed in the container before launching the server. You can also use `lmsysorg/sglang:dev-glm52-nvfp4`
+
+```sh
+pip install -U "transformers>=5.3.0" && \
+python3 -m sglang.launch_server \
+    --model nvidia/GLM-5.2-NVFP4 \
+    --tensor-parallel-size 8 \
+    --quantization modelopt_fp4 \
+    --tool-call-parser glm47 \
+    --reasoning-parser glm45 \
+    --trust-remote-code \
+    --chunked-prefill-size 16384 \
+    --mem-fraction-static 0.80
+```
+
+### vLLM
+
+To serve this checkpoint with [vLLM](https://github.com/vllm-project/vllm), use the `vllm/vllm-openai:v0.23.0` image and run:
+
+```sh
+vllm serve nvidia/GLM-5.2-NVFP4 \
+    --tensor-parallel-size 8 \
+    --enable-expert-parallel \
+    --trust-remote-code \
+    --reasoning-parser glm45 \
+    --tool-call-parser glm47 \
+    --enable-auto-tool-choice \
+    --kv-cache-dtype fp8_e4m3 \
+    --host 0.0.0.0 --port 8000
+```
+
+## Evaluation
+The accuracy benchmark results are presented in the table below. AA-LCR was measured with SGLang; all other benchmarks were measured with vLLM.
+<table>
+  <tr>
+   <td><strong>Precision</strong>
+   </td>
+   <td><strong>GPQA Diamond</strong>
+   </td>
+   <td><strong>SciCode</strong>
+   </td>
+   <td><strong>IFBench</strong>
+   </td>
+   <td><strong>AA-LCR</strong>
+   </td>
+   <td><strong>τ²-Bench Telecom</strong>
+   </td>
+  </tr>
+  <tr>
+   <td>baseline (FP8)
+   </td>
+   <td>89.52
+   </td>
+   <td>49.85
+   </td>
+   <td>74.95
+   </td>
+   <td>69.38
+   </td>
+   <td>97.9
+   </td>
+  </tr>
+  <tr>
+   <td>NVFP4
+   </td>
+   <td>89.39
+   </td>
+   <td>49.04
+   </td>
+   <td>75.81
+   </td>
+   <td>70.13
+   </td>
+   <td>98.25
+   </td>
+  </tr>
+</table>
+
+> Baseline: [GLM-5.2-FP8](https://huggingface.co/zai-org/GLM-5.2-FP8).
+> Benchmarked with temperature=1.0, top_p=0.95. GPQA Diamond used max_new_tokens=100000; all other benchmarks used max_new_tokens=64000.
+
+## Model Limitations:
+The base model was trained on data that contains toxic language and societal biases originally crawled from the internet. Therefore, the model may amplify those biases and return toxic responses especially when prompted with toxic prompts. The model may generate answers that may be inaccurate, omit key information, or include irrelevant or redundant text producing socially unacceptable or undesirable text, even if the prompt itself does not include anything explicitly offensive.
+
+## Ethical Considerations
+
+NVIDIA believes Trustworthy AI is a shared responsibility and we have established policies and practices to enable development for a wide array of AI applications. Developers should work with their internal model team to ensure this model meets requirements for the relevant industry and use case and addresses unforeseen product misuse.
+
+Please make sure you have proper rights and permissions for all input image and video content; if image or video includes people, personal health information, or intellectual property, the image or video generated will not blur or maintain proportions of image subjects included.
+
+For more detailed information on ethical considerations for this model, please see the [Model Card++ Bias, Explainability, Safety & Security, and Privacy Subcards](https://gitlab-master.nvidia.com/api-catalog/examples).
+
+Please report model quality, risk, security vulnerabilities or NVIDIA AI Concerns [here](https://www.nvidia.com/en-us/support/submit-security-vulnerability/).  <br>
+
+SUBCARDS:
+
+# **Explainability**
+
+|Field:|Response:|  
+|:---:|:---:|  
+|Intended Task/Domain:| Text generation, reasoning, summarization, and question answering. |  
+|Model Type: |Text and Image-to-text transformer |  
+|Intended Users:|This model is intended for developers, researchers, and customers building/utilizing LLMs, while balancing accuracy and efficiency.|  
+|Output:|Text String(s)|  
+|Describe how the model works:|Generates text by predicting the next word or token based on the context provided in the input sequence using multiple self-attention layers|  
+|Name the adversely impacted groups this has been tested to deliver comparable outcomes regardless of:  |  Not Applicable|
+|Technical Limitations & Mitigation:| The model was trained on data that contains toxic language and societal biases originally crawled from the internet. Therefore, the model may amplify those biases and return toxic responses especially when prompted with toxic prompts. Therefore, before deploying any applications of this model, developers should perform safety testing and tuning tailored to their specific applications of the model.|  
+|Verified to have met prescribed quality standards?|Yes|  
+|Performance Metrics:|Accuracy, Throughput, and user-side throughput|  
+|Potential Known Risk| The model may generate answers that may be inaccurate, omit key information, or include irrelevant or redundant text producing socially unacceptable or undesirable text, even if the prompt itself does not include anything explicitly offensive. |
+|Licensing:| Your usage is governed by the following **GOVERNING TERMS:** Use of the model is governed by the [MIT License](https://huggingface.co/datasets/choosealicense/licenses/blob/main/markdown/mit.md), same as the base model. |
+
+# **Bias**
+
+|Field:|Response:|  
+|:---:|:---:|  
+|Participation considerations from adversely impacted groups [protected classes](https://www.senate.ca.gov/content/protected-classes) in model design and testing:|None|  
+|Measures taken to mitigate against unwanted bias:|None|
+
+# **Safety & Security**
+
+|Field:|Response:|  
+|:---:|:---:|  
+|Model Application(s):|Chat, Instruction Following, Chatbot Development, Code Generation, Reasoning|  
+|Describe life critical application (if present):|Not Applicable |  
+|Use Case Restrictions:|Abide by the **GOVERNING TERMS:** Use of the model is governed by the [MIT License](https://huggingface.co/datasets/choosealicense/licenses/blob/main/markdown/mit.md), same as the base model. |  
+|Model and Dataset Restrictions:|The Principle of least privilege (PoLP) is applied limiting access for dataset generation.  Restrictions enforce dataset access during training, and dataset license constraints adhered to. Model checkpoints are made available on Hugging Face, and may become available on cloud providers' model catalog.|
+
+# **Privacy**
+
+|Field:|Response:|  
+|:---:|:---:|  
+|Generatable or Reverse engineerable personal data?|No|  
+|Personal data used to create this model?|No|
+|Was consent obtained for any personal data used?|Not Applicable|  
+|How often is dataset reviewed?|Before Release|  
+|Was data from user interactions with the AI model (e.g. user input and prompts) used to train the model? | No |
+|Is there provenance for all datasets used in training?|Yes|  
+|Does data labeling (annotation, metadata) comply with privacy laws?|Yes| 
+|Is data compliant with data subject requests for data correction or removal, if such a request was made? |  Not Applicable| 
+|Applicable NVIDIA Privacy Policy|https://www.nvidia.com/en-us/about-nvidia/privacy-policy/|
\ No newline at end of file
diff --git a/chat_template.jinja b/chat_template.jinja
new file mode 100644
index 0000000000000000000000000000000000000000..672034e3a1196d8b9fc0116ddc037cb554227f89
--- /dev/null
+++ b/chat_template.jinja
@@ -0,0 +1,119 @@
+[gMASK]<sop>
+{%- set effective_reasoning_effort = 'high' if reasoning_effort is defined and reasoning_effort == 'high' else 'max' -%}
+{%- if (enable_thinking is not defined or enable_thinking) and effective_reasoning_effort is not none -%}<|system|>Reasoning Effort: {{ effective_reasoning_effort | capitalize }}{%- endif -%}
+{%- if tools -%}
+{%- macro tool_to_json(tool) -%}
+    {%- set ns_tool = namespace(first=true) -%}
+    {{ '{' -}}
+    {%- for k, v in tool.items() -%}
+        {%- if k != 'defer_loading' and k != 'strict' -%}
+            {%- if not ns_tool.first -%}{{- ', ' -}}{%- endif -%}
+            {%- set ns_tool.first = false -%}
+            "{{ k }}": {{ v | tojson(ensure_ascii=False) }}
+        {%- endif -%}
+    {%- endfor -%}
+    {{- '}' -}}
+{%- endmacro -%}
+<|system|>
+# Tools
+
+You may call one or more functions to assist with the user query.
+
+You are provided with function signatures within <tools></tools> XML tags:
+<tools>
+{% for tool in tools %}
+{%- if 'function' in tool -%}
+    {%- set tool = tool['function'] -%}
+{%- endif -%}
+{% if tool.defer_loading is not defined or not tool.defer_loading %}
+{{ tool_to_json(tool) }}
+{% endif %}
+{% endfor %}
+</tools>
+
+For each function call, output the function name and arguments within the following XML format:
+<tool_call>{function-name}<arg_key>{arg-key-1}</arg_key><arg_value>{arg-value-1}</arg_value><arg_key>{arg-key-2}</arg_key><arg_value>{arg-value-2}</arg_value>...</tool_call>{%- endif -%}
+{%- macro visible_text(content) -%}
+    {%- if content is string -%}
+        {{- content }}
+    {%- elif content is iterable and content is not mapping -%}
+        {%- for item in content -%}
+            {%- if item is mapping and item.type == 'text' -%}
+                {{- item.text }}
+            {%- elif item is string -%}
+                {{- item }}
+            {%- elif item is mapping and item.type in ['image', 'image_url', 'video', 'video_url', 'audio', 'audio_url', 'input_audio'] -%}
+                {%- set media_type = item.type | replace('_url', '') | replace('input_', '') -%}
+                {{- "<reminder>You are unable to process this " ~ media_type ~ " because you don't have multi-modal input ability. Try different methods.</reminder>" }}
+            {%- endif -%}
+        {%- endfor -%}
+    {%- else -%}
+        {{- content }}
+    {%- endif -%}
+{%- endmacro -%}
+{%- set ns = namespace(last_user_index=-1) -%}
+{%- for m in messages %}
+    {%- if m.role == 'user' %}
+        {%- set ns.last_user_index = loop.index0 -%}
+    {%- endif %}
+{%- endfor %}
+{%- for m in messages -%}
+{%- if m.role == 'user' -%}<|user|>{{ visible_text(m.content) }}
+{%- elif m.role == 'assistant' -%}
+<|assistant|>
+{%- set content = visible_text(m.content) %}
+{%- if m.reasoning_content is string %}
+    {%- set reasoning_content = m.reasoning_content %}
+{%- elif '</think>' in content %}
+    {%- set reasoning_content = content.split('</think>')[0].split('<think>')[-1] %}
+    {%- set content = content.split('</think>')[-1] %}
+{%- endif %}
+{%- if ((clear_thinking is defined and not clear_thinking) or loop.index0 > ns.last_user_index) and reasoning_content is defined -%}
+{{ '<think>' + reasoning_content +  '</think>'}}
+{%- else -%}
+{{ '<think></think>' }}
+{%- endif -%}
+{%- if content.strip() -%}
+{{ content.strip() }}
+{%- endif -%}
+{% if m.tool_calls %}
+{% for tc in m.tool_calls %}
+{%- if tc.function %}
+    {%- set tc = tc.function %}
+{%- endif %}
+{{- '<tool_call>' + tc.name -}}
+{% set _args = tc.arguments %}{% for k, v in _args.items() %}<arg_key>{{ k }}</arg_key><arg_value>{{ v | tojson(ensure_ascii=False) if v is not string else v }}</arg_value>{% endfor %}</tool_call>{% endfor %}
+{% endif %}
+{%- elif m.role == 'tool' -%}
+{%- if loop.first or (messages[loop.index0 - 1].role != "tool") %}
+    {{- '<|observation|>' -}}
+{%- endif %}
+{%- if m.content is string -%}
+    {{- '<tool_response>' + m.content + '</tool_response>' -}}
+{%- elif m.content is iterable and m.content is not mapping and m.content and m.content.0.type == "tool_reference" -%}
+    {{- '<tool_response><tools>\n' -}}
+    {% for tr in m.content %}
+        {%- for tool in tools -%}
+            {%- if 'function' in tool -%}
+                {%- set tool = tool['function'] -%}
+            {%- endif -%}
+            {%- if tool.name == tr.name -%}
+                {{- tool_to_json(tool) + '\n' -}}
+            {%- endif -%}
+        {%- endfor -%}
+    {%- endfor -%}
+    {{- '</tools></tool_response>' -}}
+{%- elif m.content is iterable and m.content is not mapping and m.content and m.content.0 is mapping and m.content.0.output is defined -%}
+    {%- for tr in m.content -%}
+        {{- '<tool_response>' + tr.output + '</tool_response>' -}}
+    {%- endfor -%}
+{%- else -%}
+    {{- '<tool_response>' + visible_text(m.content) + '</tool_response>' -}}
+{% endif -%}
+{%- elif m.role == 'system' -%}
+<|system|>{{ visible_text(m.content) }}
+{%- endif -%}
+{%- endfor -%}
+{%- if add_generation_prompt -%}
+    <|assistant|>{{- '<think></think>' if (enable_thinking is defined and not enable_thinking) else '<think>' -}}
+{%- endif -%}
diff --git a/config.json b/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..b63d4c44cdd013e928cbc37a755633e8c001626f
--- /dev/null
+++ b/config.json
@@ -0,0 +1,496 @@
+{
+    "architectures": [
+        "GlmMoeDsaForCausalLM"
+    ],
+    "attention_bias": false,
+    "attention_dropout": 0.0,
+    "bos_token_id": 0,
+    "dtype": "bfloat16",
+    "eos_token_id": [
+        154820,
+        154827,
+        154829
+    ],
+    "ep_size": 1,
+    "first_k_dense_replace": 3,
+    "head_dim": 192,
+    "hidden_act": "silu",
+    "hidden_size": 6144,
+    "index_head_dim": 128,
+    "index_n_heads": 32,
+    "index_share_for_mtp_iteration": true,
+    "index_skip_topk_offset": 3,
+    "index_topk": 2048,
+    "index_topk_freq": 4,
+    "index_topk_pattern": null,
+    "indexer_rope_interleave": true,
+    "indexer_types": [
+        "full",
+        "full",
+        "full",
+        "shared",
+        "shared",
+        "shared",
+        "full",
+        "shared",
+        "shared",
+        "shared",
+        "full",
+        "shared",
+        "shared",
+        "shared",
+        "full",
+        "shared",
+        "shared",
+        "shared",
+        "full",
+        "shared",
+        "shared",
+        "shared",
+        "full",
+        "shared",
+        "shared",
+        "shared",
+        "full",
+        "shared",
+        "shared",
+        "shared",
+        "full",
+        "shared",
+        "shared",
+        "shared",
+        "full",
+        "shared",
+        "shared",
+        "shared",
+        "full",
+        "shared",
+        "shared",
+        "shared",
+        "full",
+        "shared",
+        "shared",
+        "shared",
+        "full",
+        "shared",
+        "shared",
+        "shared",
+        "full",
+        "shared",
+        "shared",
+        "shared",
+        "full",
+        "shared",
+        "shared",
+        "shared",
+        "full",
+        "shared",
+        "shared",
+        "shared",
+        "full",
+        "shared",
+        "shared",
+        "shared",
+        "full",
+        "shared",
+        "shared",
+        "shared",
+        "full",
+        "shared",
+        "shared",
+        "shared",
+        "full",
+        "shared",
+        "shared",
+        "shared"
+    ],
+    "initializer_range": 0.02,
+    "intermediate_size": 12288,
+    "kv_lora_rank": 512,
+    "layer_types": [
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention",
+        "deepseek_sparse_attention"
+    ],
+    "max_position_embeddings": 1048576,
+    "mlp_bias": false,
+    "mlp_layer_types": [
+        "dense",
+        "dense",
+        "dense",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse",
+        "sparse"
+    ],
+    "model_type": "glm_moe_dsa",
+    "moe_intermediate_size": 2048,
+    "moe_layer_freq": 1,
+    "n_group": 1,
+    "n_routed_experts": 256,
+    "n_shared_experts": 1,
+    "norm_topk_prob": true,
+    "num_attention_heads": 64,
+    "num_experts": 256,
+    "num_experts_per_tok": 8,
+    "num_hidden_layers": 78,
+    "num_key_value_heads": 64,
+    "num_nextn_predict_layers": 1,
+    "pad_token_id": 154820,
+    "pretraining_tp": 1,
+    "q_lora_rank": 2048,
+    "qk_head_dim": 256,
+    "qk_nope_head_dim": 192,
+    "qk_rope_head_dim": 64,
+    "rms_norm_eps": 1e-05,
+    "rope_interleave": true,
+    "rope_parameters": {
+        "rope_theta": 8000000,
+        "rope_type": "default"
+    },
+    "routed_scaling_factor": 2.5,
+    "scoring_func": "sigmoid",
+    "tie_word_embeddings": false,
+    "topk_group": 1,
+    "topk_method": "noaux_tc",
+    "transformers_version": "5.11.0",
+    "use_cache": true,
+    "v_head_dim": 256,
+    "vocab_size": 154880,
+    "quantization_config": {
+        "config_groups": {
+            "group_0": {
+                "input_activations": {
+                    "dynamic": false,
+                    "num_bits": 4,
+                    "type": "float",
+                    "group_size": 16
+                },
+                "weights": {
+                    "dynamic": false,
+                    "num_bits": 4,
+                    "type": "float",
+                    "group_size": 16
+                },
+                "targets": [
+                    "Linear"
+                ]
+            }
+        },
+        "ignore": [
+            "lm_head",
+            "model.embed_tokens",
+            "model.layers.0*",
+            "model.layers.1.*",
+            "model.layers.10.mlp.shared_experts*",
+            "model.layers.10.self_attn*",
+            "model.layers.11.mlp.shared_experts*",
+            "model.layers.11.self_attn*",
+            "model.layers.12.mlp.shared_experts*",
+            "model.layers.12.self_attn*",
+            "model.layers.13.mlp.shared_experts*",
+            "model.layers.13.self_attn*",
+            "model.layers.14.mlp.shared_experts*",
+            "model.layers.14.self_attn*",
+            "model.layers.15.mlp.shared_experts*",
+            "model.layers.15.self_attn*",
+            "model.layers.16.mlp.shared_experts*",
+            "model.layers.16.self_attn*",
+            "model.layers.17.mlp.shared_experts*",
+            "model.layers.17.self_attn*",
+            "model.layers.18.mlp.shared_experts*",
+            "model.layers.18.self_attn*",
+            "model.layers.19.mlp.shared_experts*",
+            "model.layers.19.self_attn*",
+            "model.layers.2.*",
+            "model.layers.20.mlp.shared_experts*",
+            "model.layers.20.self_attn*",
+            "model.layers.21.mlp.shared_experts*",
+            "model.layers.21.self_attn*",
+            "model.layers.22.mlp.shared_experts*",
+            "model.layers.22.self_attn*",
+            "model.layers.23.mlp.shared_experts*",
+            "model.layers.23.self_attn*",
+            "model.layers.24.mlp.shared_experts*",
+            "model.layers.24.self_attn*",
+            "model.layers.25.mlp.shared_experts*",
+            "model.layers.25.self_attn*",
+            "model.layers.26.mlp.shared_experts*",
+            "model.layers.26.self_attn*",
+            "model.layers.27.mlp.shared_experts*",
+            "model.layers.27.self_attn*",
+            "model.layers.28.mlp.shared_experts*",
+            "model.layers.28.self_attn*",
+            "model.layers.29.mlp.shared_experts*",
+            "model.layers.29.self_attn*",
+            "model.layers.3.mlp.shared_experts*",
+            "model.layers.3.self_attn*",
+            "model.layers.30.mlp.shared_experts*",
+            "model.layers.30.self_attn*",
+            "model.layers.31.mlp.shared_experts*",
+            "model.layers.31.self_attn*",
+            "model.layers.32.mlp.shared_experts*",
+            "model.layers.32.self_attn*",
+            "model.layers.33.mlp.shared_experts*",
+            "model.layers.33.self_attn*",
+            "model.layers.34.mlp.shared_experts*",
+            "model.layers.34.self_attn*",
+            "model.layers.35.mlp.shared_experts*",
+            "model.layers.35.self_attn*",
+            "model.layers.36.mlp.shared_experts*",
+            "model.layers.36.self_attn*",
+            "model.layers.37.mlp.shared_experts*",
+            "model.layers.37.self_attn*",
+            "model.layers.38.mlp.shared_experts*",
+            "model.layers.38.self_attn*",
+            "model.layers.39.mlp.shared_experts*",
+            "model.layers.39.self_attn*",
+            "model.layers.4.mlp.shared_experts*",
+            "model.layers.4.self_attn*",
+            "model.layers.40.mlp.shared_experts*",
+            "model.layers.40.self_attn*",
+            "model.layers.41.mlp.shared_experts*",
+            "model.layers.41.self_attn*",
+            "model.layers.42.mlp.shared_experts*",
+            "model.layers.42.self_attn*",
+            "model.layers.43.mlp.shared_experts*",
+            "model.layers.43.self_attn*",
+            "model.layers.44.mlp.shared_experts*",
+            "model.layers.44.self_attn*",
+            "model.layers.45.mlp.shared_experts*",
+            "model.layers.45.self_attn*",
+            "model.layers.46.mlp.shared_experts*",
+            "model.layers.46.self_attn*",
+            "model.layers.47.mlp.shared_experts*",
+            "model.layers.47.self_attn*",
+            "model.layers.48.mlp.shared_experts*",
+            "model.layers.48.self_attn*",
+            "model.layers.49.mlp.shared_experts*",
+            "model.layers.49.self_attn*",
+            "model.layers.5.mlp.shared_experts*",
+            "model.layers.5.self_attn*",
+            "model.layers.50.mlp.shared_experts*",
+            "model.layers.50.self_attn*",
+            "model.layers.51.mlp.shared_experts*",
+            "model.layers.51.self_attn*",
+            "model.layers.52.mlp.shared_experts*",
+            "model.layers.52.self_attn*",
+            "model.layers.53.mlp.shared_experts*",
+            "model.layers.53.self_attn*",
+            "model.layers.54.mlp.shared_experts*",
+            "model.layers.54.self_attn*",
+            "model.layers.55.mlp.shared_experts*",
+            "model.layers.55.self_attn*",
+            "model.layers.56.mlp.shared_experts*",
+            "model.layers.56.self_attn*",
+            "model.layers.57.mlp.shared_experts*",
+            "model.layers.57.self_attn*",
+            "model.layers.58.mlp.shared_experts*",
+            "model.layers.58.self_attn*",
+            "model.layers.59.mlp.shared_experts*",
+            "model.layers.59.self_attn*",
+            "model.layers.6.mlp.shared_experts*",
+            "model.layers.6.self_attn*",
+            "model.layers.60.mlp.shared_experts*",
+            "model.layers.60.self_attn*",
+            "model.layers.61.mlp.shared_experts*",
+            "model.layers.61.self_attn*",
+            "model.layers.62.mlp.shared_experts*",
+            "model.layers.62.self_attn*",
+            "model.layers.63.mlp.shared_experts*",
+            "model.layers.63.self_attn*",
+            "model.layers.64.mlp.shared_experts*",
+            "model.layers.64.self_attn*",
+            "model.layers.65.mlp.shared_experts*",
+            "model.layers.65.self_attn*",
+            "model.layers.66.mlp.shared_experts*",
+            "model.layers.66.self_attn*",
+            "model.layers.67.mlp.shared_experts*",
+            "model.layers.67.self_attn*",
+            "model.layers.68.mlp.shared_experts*",
+            "model.layers.68.self_attn*",
+            "model.layers.69.mlp.shared_experts*",
+            "model.layers.69.self_attn*",
+            "model.layers.7.mlp.shared_experts*",
+            "model.layers.7.self_attn*",
+            "model.layers.70.mlp.shared_experts*",
+            "model.layers.70.self_attn*",
+            "model.layers.71.mlp.shared_experts*",
+            "model.layers.71.self_attn*",
+            "model.layers.72.mlp.shared_experts*",
+            "model.layers.72.self_attn*",
+            "model.layers.73.mlp.shared_experts*",
+            "model.layers.73.self_attn*",
+            "model.layers.74.mlp.shared_experts*",
+            "model.layers.74.self_attn*",
+            "model.layers.75.mlp.shared_experts*",
+            "model.layers.75.self_attn*",
+            "model.layers.76.mlp.shared_experts*",
+            "model.layers.76.self_attn*",
+            "model.layers.77.mlp.shared_experts*",
+            "model.layers.77.self_attn*",
+            "model.layers.8.mlp.shared_experts*",
+            "model.layers.8.self_attn*",
+            "model.layers.9.mlp.shared_experts*",
+            "model.layers.9.self_attn*",
+            "model.layers.78*"
+        ],
+        "quant_algo": "NVFP4",
+        "kv_cache_scheme": {
+            "dynamic": false,
+            "num_bits": 8,
+            "type": "float"
+        },
+        "producer": {
+            "name": "modelopt",
+            "version": "0.46.0.dev65+g977d34dc3"
+        },
+        "quant_method": "modelopt"
+    }
+}
diff --git a/generation_config.json b/generation_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..4780e3aba8759054a8f8b3e4d1dc2277c1353e14
--- /dev/null
+++ b/generation_config.json
@@ -0,0 +1,13 @@
+{
+  "_from_model_config": true,
+  "do_sample": true,
+  "eos_token_id": [
+    154820,
+    154827,
+    154829
+  ],
+  "pad_token_id": 154820,
+  "temperature": 1.0,
+  "top_p": 0.95,
+  "transformers_version": "5.11.0"
+}
diff --git a/hf_quant_config.json b/hf_quant_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..4c1380a3df73173f8662564924705089a6220958
--- /dev/null
+++ b/hf_quant_config.json
@@ -0,0 +1,169 @@
+{
+    "producer": {
+        "name": "modelopt",
+        "version": "0.46.0.dev65+g977d34dc3"
+    },
+    "quantization": {
+        "quant_algo": "NVFP4",
+        "kv_cache_quant_algo": "FP8",
+        "group_size": 16,
+        "exclude_modules": [
+            "lm_head",
+            "model.embed_tokens",
+            "model.layers.0*",
+            "model.layers.1.*",
+            "model.layers.10.mlp.shared_experts*",
+            "model.layers.10.self_attn*",
+            "model.layers.11.mlp.shared_experts*",
+            "model.layers.11.self_attn*",
+            "model.layers.12.mlp.shared_experts*",
+            "model.layers.12.self_attn*",
+            "model.layers.13.mlp.shared_experts*",
+            "model.layers.13.self_attn*",
+            "model.layers.14.mlp.shared_experts*",
+            "model.layers.14.self_attn*",
+            "model.layers.15.mlp.shared_experts*",
+            "model.layers.15.self_attn*",
+            "model.layers.16.mlp.shared_experts*",
+            "model.layers.16.self_attn*",
+            "model.layers.17.mlp.shared_experts*",
+            "model.layers.17.self_attn*",
+            "model.layers.18.mlp.shared_experts*",
+            "model.layers.18.self_attn*",
+            "model.layers.19.mlp.shared_experts*",
+            "model.layers.19.self_attn*",
+            "model.layers.2.*",
+            "model.layers.20.mlp.shared_experts*",
+            "model.layers.20.self_attn*",
+            "model.layers.21.mlp.shared_experts*",
+            "model.layers.21.self_attn*",
+            "model.layers.22.mlp.shared_experts*",
+            "model.layers.22.self_attn*",
+            "model.layers.23.mlp.shared_experts*",
+            "model.layers.23.self_attn*",
+            "model.layers.24.mlp.shared_experts*",
+            "model.layers.24.self_attn*",
+            "model.layers.25.mlp.shared_experts*",
+            "model.layers.25.self_attn*",
+            "model.layers.26.mlp.shared_experts*",
+            "model.layers.26.self_attn*",
+            "model.layers.27.mlp.shared_experts*",
+            "model.layers.27.self_attn*",
+            "model.layers.28.mlp.shared_experts*",
+            "model.layers.28.self_attn*",
+            "model.layers.29.mlp.shared_experts*",
+            "model.layers.29.self_attn*",
+            "model.layers.3.mlp.shared_experts*",
+            "model.layers.3.self_attn*",
+            "model.layers.30.mlp.shared_experts*",
+            "model.layers.30.self_attn*",
+            "model.layers.31.mlp.shared_experts*",
+            "model.layers.31.self_attn*",
+            "model.layers.32.mlp.shared_experts*",
+            "model.layers.32.self_attn*",
+            "model.layers.33.mlp.shared_experts*",
+            "model.layers.33.self_attn*",
+            "model.layers.34.mlp.shared_experts*",
+            "model.layers.34.self_attn*",
+            "model.layers.35.mlp.shared_experts*",
+            "model.layers.35.self_attn*",
+            "model.layers.36.mlp.shared_experts*",
+            "model.layers.36.self_attn*",
+            "model.layers.37.mlp.shared_experts*",
+            "model.layers.37.self_attn*",
+            "model.layers.38.mlp.shared_experts*",
+            "model.layers.38.self_attn*",
+            "model.layers.39.mlp.shared_experts*",
+            "model.layers.39.self_attn*",
+            "model.layers.4.mlp.shared_experts*",
+            "model.layers.4.self_attn*",
+            "model.layers.40.mlp.shared_experts*",
+            "model.layers.40.self_attn*",
+            "model.layers.41.mlp.shared_experts*",
+            "model.layers.41.self_attn*",
+            "model.layers.42.mlp.shared_experts*",
+            "model.layers.42.self_attn*",
+            "model.layers.43.mlp.shared_experts*",
+            "model.layers.43.self_attn*",
+            "model.layers.44.mlp.shared_experts*",
+            "model.layers.44.self_attn*",
+            "model.layers.45.mlp.shared_experts*",
+            "model.layers.45.self_attn*",
+            "model.layers.46.mlp.shared_experts*",
+            "model.layers.46.self_attn*",
+            "model.layers.47.mlp.shared_experts*",
+            "model.layers.47.self_attn*",
+            "model.layers.48.mlp.shared_experts*",
+            "model.layers.48.self_attn*",
+            "model.layers.49.mlp.shared_experts*",
+            "model.layers.49.self_attn*",
+            "model.layers.5.mlp.shared_experts*",
+            "model.layers.5.self_attn*",
+            "model.layers.50.mlp.shared_experts*",
+            "model.layers.50.self_attn*",
+            "model.layers.51.mlp.shared_experts*",
+            "model.layers.51.self_attn*",
+            "model.layers.52.mlp.shared_experts*",
+            "model.layers.52.self_attn*",
+            "model.layers.53.mlp.shared_experts*",
+            "model.layers.53.self_attn*",
+            "model.layers.54.mlp.shared_experts*",
+            "model.layers.54.self_attn*",
+            "model.layers.55.mlp.shared_experts*",
+            "model.layers.55.self_attn*",
+            "model.layers.56.mlp.shared_experts*",
+            "model.layers.56.self_attn*",
+            "model.layers.57.mlp.shared_experts*",
+            "model.layers.57.self_attn*",
+            "model.layers.58.mlp.shared_experts*",
+            "model.layers.58.self_attn*",
+            "model.layers.59.mlp.shared_experts*",
+            "model.layers.59.self_attn*",
+            "model.layers.6.mlp.shared_experts*",
+            "model.layers.6.self_attn*",
+            "model.layers.60.mlp.shared_experts*",
+            "model.layers.60.self_attn*",
+            "model.layers.61.mlp.shared_experts*",
+            "model.layers.61.self_attn*",
+            "model.layers.62.mlp.shared_experts*",
+            "model.layers.62.self_attn*",
+            "model.layers.63.mlp.shared_experts*",
+            "model.layers.63.self_attn*",
+            "model.layers.64.mlp.shared_experts*",
+            "model.layers.64.self_attn*",
+            "model.layers.65.mlp.shared_experts*",
+            "model.layers.65.self_attn*",
+            "model.layers.66.mlp.shared_experts*",
+            "model.layers.66.self_attn*",
+            "model.layers.67.mlp.shared_experts*",
+            "model.layers.67.self_attn*",
+            "model.layers.68.mlp.shared_experts*",
+            "model.layers.68.self_attn*",
+            "model.layers.69.mlp.shared_experts*",
+            "model.layers.69.self_attn*",
+            "model.layers.7.mlp.shared_experts*",
+            "model.layers.7.self_attn*",
+            "model.layers.70.mlp.shared_experts*",
+            "model.layers.70.self_attn*",
+            "model.layers.71.mlp.shared_experts*",
+            "model.layers.71.self_attn*",
+            "model.layers.72.mlp.shared_experts*",
+            "model.layers.72.self_attn*",
+            "model.layers.73.mlp.shared_experts*",
+            "model.layers.73.self_attn*",
+            "model.layers.74.mlp.shared_experts*",
+            "model.layers.74.self_attn*",
+            "model.layers.75.mlp.shared_experts*",
+            "model.layers.75.self_attn*",
+            "model.layers.76.mlp.shared_experts*",
+            "model.layers.76.self_attn*",
+            "model.layers.77.mlp.shared_experts*",
+            "model.layers.77.self_attn*",
+            "model.layers.8.mlp.shared_experts*",
+            "model.layers.8.self_attn*",
+            "model.layers.9.mlp.shared_experts*",
+            "model.layers.9.self_attn*",
+            "model.layers.78*"
+        ]
+    }
+}
\ No newline at end of file
diff --git a/model-00001-of-00047.safetensors b/model-00001-of-00047.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..242cebfe78136605b24e9ea246ec3d3d5ec0efdf
--- /dev/null
+++ b/model-00001-of-00047.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7355ca8065dac64f806bb8f51921e9c86f428542815e4cbfbe67c207a9e0ff4d
+size 9996935800
diff --git a/model-00002-of-00047.safetensors b/model-00002-of-00047.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..8bd336d24af3e761db4782e2c5381867531646c9
--- /dev/null
+++ b/model-00002-of-00047.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5a93bf6840c92c2b921a696b3268c3583f417ce4b85fdf912a3b164679d1df9b
+size 9998076664
diff --git a/model-00003-of-00047.safetensors b/model-00003-of-00047.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..ecd0504870835888ab4e8ead23fdec1356d1530e
--- /dev/null
+++ b/model-00003-of-00047.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:240485288e04efae1c4dcda20eb13a4376efc5f69c0ea1b8c5e3d5d771707ea1
+size 9995585544
diff --git a/model-00004-of-00047.safetensors b/model-00004-of-00047.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..92759a593a9dd5975548fb324f9b4b591da3cb57
--- /dev/null
+++ b/model-00004-of-00047.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:790fe046fb8b18f1c1bc829c32ae88b02aba7203a8fbdb1acf864b37dc838fae
+size 9998076664
diff --git a/model-00005-of-00047.safetensors b/model-00005-of-00047.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..3c02a65c8ca5ec9bc0a0af0ffb2a34f9e63c4342
--- /dev/null
+++ b/model-00005-of-00047.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9c6a3381b16ff8f37bed8e6e020b338b569f1c927b9a3fa8c025c5c75c77ca79
+size 9997417192
diff --git a/model-00006-of-00047.safetensors b/model-00006-of-00047.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..55dbca47e1cab0c7e9bc47162d07e40874147a38
--- /dev/null
+++ b/model-00006-of-00047.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:902e254158f0fac8bb531ff85728de4c1580272464a3418d1e452d67ab9ee4d0
+size 9998081432
diff --git a/model-00007-of-00047.safetensors b/model-00007-of-00047.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..6c3dd7094bd55d3de8dfe71cb24b59d6ef0ab5f2
--- /dev/null
+++ b/model-00007-of-00047.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:40299690b69637fc15fddf048b47ef0bc82a3c6d9685f0fd048a9823b426cdae
+size 9995590872
diff --git a/model-00008-of-00047.safetensors b/model-00008-of-00047.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..c1ecfe8806a670b51b794ea7a96ab213505a61a5
--- /dev/null
+++ b/model-00008-of-00047.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:139f1c107077c04c2646d60d16f2c723d8be1876efa8e271eb135f4e564fdb31
+size 9999911744
diff --git a/model-00009-of-00047.safetensors b/model-00009-of-00047.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..02ba9c00fcf0534348f0956f734e53cc1cb410dc
--- /dev/null
+++ b/model-00009-of-00047.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e1db5be0dac72aaaa446e1fa2437420cf9e47efe3c36c59c5471200be1dd4d23
+size 9998081760
diff --git a/model-00010-of-00047.safetensors b/model-00010-of-00047.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..b4825695a76ef8eb5e86947f3673798d4b2f06ed
--- /dev/null
+++ b/model-00010-of-00047.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:617a2ca6fd37281f636ed819eab39dd8f2ebb21bc8bbf599f229537bdadb2024
+size 9995590752
diff --git a/model-00011-of-00047.safetensors b/model-00011-of-00047.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..075ab4f39a637e74bfc0820e558900c01883cff2
--- /dev/null
+++ b/model-00011-of-00047.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5099883c548f807bc1f2cde6188425ac6f40a9c3c244a4e18bfc008e572e7361
+size 9998081968
diff --git a/model-00012-of-00047.safetensors b/model-00012-of-00047.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..da4b319543bbf5457d574452503cad80261aa92a
--- /dev/null
+++ b/model-00012-of-00047.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:df6a0c6ab9fa7be42f0db0872880f1b5636e910b4f73aac5570c42354f05804c
+size 9997420424
diff --git a/model-00013-of-00047.safetensors b/model-00013-of-00047.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..29b3ce8a47cb9b7f0408cb54db27f0a18268b142
--- /dev/null
+++ b/model-00013-of-00047.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2bfd1392eda129e70128ef242284219e5acd07f3b77eeea9f7a6e6b48c3020b9
+size 9998081432
diff --git a/model-00014-of-00047.safetensors b/model-00014-of-00047.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..4604d1fccea5c2aabb73cc05c10126bf145ead25
--- /dev/null
+++ b/model-00014-of-00047.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6320d889ab91a2ba274bba38b8ea19b7540b20a6b484ec522ca8bd96ea5333ca
+size 9995590992
diff --git a/model-00015-of-00047.safetensors b/model-00015-of-00047.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..fe3fabba31426f578b48ba716174982da52be392
--- /dev/null
+++ b/model-00015-of-00047.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:499d121db3e7e08e403a1665071bf55f710d88916de32519359691a571bcbb09
+size 9999911624
diff --git a/model-00016-of-00047.safetensors b/model-00016-of-00047.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..8a83d6ee8d390585f4f00ce44b736a292f14f0cc
--- /dev/null
+++ b/model-00016-of-00047.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b0bfeb109e4a1ca9327430de28f11e858d239505a4d68cc05f93440818c3352b
+size 9998081640
diff --git a/model-00017-of-00047.safetensors b/model-00017-of-00047.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..f8af604ba5dd1b06d5b2fdee63048132548482d6
--- /dev/null
+++ b/model-00017-of-00047.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ca88817f6f027fec523052c67e08de81a3751b3b44b45026a29321f32ba61b4b
+size 9995590752
diff --git a/model-00018-of-00047.safetensors b/model-00018-of-00047.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..04849012c513288765e703b04a76d600cc664b70
--- /dev/null
+++ b/model-00018-of-00047.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4d32422c750fe5520879ce63b037d2c7061abd0975c56011f757eea603bfbf4b
+size 9959557432
diff --git a/model-00019-of-00047.safetensors b/model-00019-of-00047.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..bd31b5c2edf6186781774fab0fae61bc0a20b3aa
--- /dev/null
+++ b/model-00019-of-00047.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:74ecfce9cddd66f2f6ef551c6f485c1eb2ce5f9f63a9953773c2ab50ce2eff68
+size 10000552960
diff --git a/model-00020-of-00047.safetensors b/model-00020-of-00047.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..bc745a077a358594904f5e484f5f47d2e26cb38e
--- /dev/null
+++ b/model-00020-of-00047.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:23a23304160cf55ca305e920af3f47aabbfacb6a7eb3c2d6624e3cd45af79656
+size 9998081424
diff --git a/model-00021-of-00047.safetensors b/model-00021-of-00047.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..db885304310b28e9c2c0268bbc4b9439e128c829
--- /dev/null
+++ b/model-00021-of-00047.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:db14cdfbe3e53abbd291c230c148d9c21fd9834f1a6e6bdde8d860936a91c382
+size 9995591144
diff --git a/model-00022-of-00047.safetensors b/model-00022-of-00047.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..e03549cf940fcbbd69e1b0c10a77289c4d8a3166
--- /dev/null
+++ b/model-00022-of-00047.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bf5f1c293e95bc3f37f9da5b2c464d765b9150006695ef4b76d973c2fdc1ef75
+size 9999911480
diff --git a/model-00023-of-00047.safetensors b/model-00023-of-00047.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..c9858a69a6abfaab70b24cd8270f0291e0328500
--- /dev/null
+++ b/model-00023-of-00047.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2fb70768ccd149d6df403ff319b831d3578c6c7ea7040bd6350d36201e2325ae
+size 9998081504
diff --git a/model-00024-of-00047.safetensors b/model-00024-of-00047.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..c0a4e456d4a5eeb8a9852a05d8d75915163bd926
--- /dev/null
+++ b/model-00024-of-00047.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ccf3b2e4ca1e8badbb7e0e3b5e7ed2f23a11a996ee35496f60748649ca45577e
+size 9995590752
diff --git a/model-00025-of-00047.safetensors b/model-00025-of-00047.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..da26667db36697f683c5da5909e2ef211186e3ba
--- /dev/null
+++ b/model-00025-of-00047.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:67f233905c40eaa27e1d9174cefd28403ea55dd195f3f97dfa52b55eb480572c
+size 9999911904
diff --git a/model-00026-of-00047.safetensors b/model-00026-of-00047.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..ab4a76af87abfb4ecc4dadc3cdcb15ad6f7e81dc
--- /dev/null
+++ b/model-00026-of-00047.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:daf23e3ccbf4b5324bcc77a208b4f91f1ab4965a24df62524e26e19a2fef2585
+size 9995591304
diff --git a/model-00027-of-00047.safetensors b/model-00027-of-00047.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..8a53599068262afbb93db1133d9d714105c54aad
--- /dev/null
+++ b/model-00027-of-00047.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1b4c7404684d1f71e6341660ce60b87184e9295c042c68816f70cf21e0d5d378
+size 9998081416
diff --git a/model-00028-of-00047.safetensors b/model-00028-of-00047.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..44fd77071e27c78413889fed2bda1dd6fe7d08af
--- /dev/null
+++ b/model-00028-of-00047.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a895f3e5a8150c74996caf4fe7c8b7f11f4ae99f838bcc3207af1a9f87010199
+size 9995591208
diff --git a/model-00029-of-00047.safetensors b/model-00029-of-00047.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..f7c008ac73b178d2eced7a59b3279d53ae5e668e
--- /dev/null
+++ b/model-00029-of-00047.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5601bf446f5f778b3c2a26d023c78d21c0ed3200c1f2c1a13b082a94143937f5
+size 9999911360
diff --git a/model-00030-of-00047.safetensors b/model-00030-of-00047.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..a79cfecc44a81aead4dce18a7149a5b7b8deac6d
--- /dev/null
+++ b/model-00030-of-00047.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:355b7b4665a899fad5b5b42903c8ff8f58cee469883a7638e651f8ab14a92487
+size 9998081432
diff --git a/model-00031-of-00047.safetensors b/model-00031-of-00047.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..68b3955d036c1e0328045e6a2e71ac3403ab8e56
--- /dev/null
+++ b/model-00031-of-00047.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7bc1ca71bffdd1a1eee33e43a328d525e23cf39db1b4812fdf55dd3cd90731a4
+size 9995590840
diff --git a/model-00032-of-00047.safetensors b/model-00032-of-00047.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..6f8d22211759df282ba741483e071ad0ce646b70
--- /dev/null
+++ b/model-00032-of-00047.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d11bd3fe6f4c772a754ab4b19eca4e05c231c610d2f565aae9f1865723cfef10
+size 9999911784
diff --git a/model-00033-of-00047.safetensors b/model-00033-of-00047.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..c4fe91ef17e48165331eee921220f42a879f0330
--- /dev/null
+++ b/model-00033-of-00047.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e39c9ff0d42c6052c538aaee35579a703993862601a340d0c4066ef0d2806bbd
+size 9995591128
diff --git a/model-00034-of-00047.safetensors b/model-00034-of-00047.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..af6b37ec90a4e3eff0f9065eed3c52f95264983d
--- /dev/null
+++ b/model-00034-of-00047.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a81b529612c93893cf55dfb9c1f1a9a0ebdcca494f62d66181d9c6577a8ed434
+size 9998081416
diff --git a/model-00035-of-00047.safetensors b/model-00035-of-00047.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..d3bcc68a80f88f250dfaf3a7471fd14c56831c74
--- /dev/null
+++ b/model-00035-of-00047.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:70d19f0b6180627f5da59d38ca2d944b3937b77ba0302a5d60db7bfa7a69093c
+size 9995591272
diff --git a/model-00036-of-00047.safetensors b/model-00036-of-00047.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..86ac681c8db44124aed6b6c64326fa9dae83b3ed
--- /dev/null
+++ b/model-00036-of-00047.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cab4d531d01434f35ec2809ad408519f52b733dd46f30adc43f3aaef6ce95933
+size 9999911176
diff --git a/model-00037-of-00047.safetensors b/model-00037-of-00047.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..5cc05f40c7048187a30cc59ffed04f0f1971fcfb
--- /dev/null
+++ b/model-00037-of-00047.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dc213d3550fc4e8482f88e08c791b4ba74fa8bb85ec2c6d7fc15661dac954055
+size 9998081432
diff --git a/model-00038-of-00047.safetensors b/model-00038-of-00047.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..2bf70a2ae1b534b0c1086302f70c931f3068fac6
--- /dev/null
+++ b/model-00038-of-00047.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:50c4a5522bd9e4a34977f52fe55018812f483a2afc8c7a32c92cc94e268b474a
+size 9995590960
diff --git a/model-00039-of-00047.safetensors b/model-00039-of-00047.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..97064d37d8111c3c74801a3152f364443bbc82be
--- /dev/null
+++ b/model-00039-of-00047.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ca17fd22173184cc0dd50977538725c11f2e6a44d65a48865dc8ea975cfffa67
+size 9999911664
diff --git a/model-00040-of-00047.safetensors b/model-00040-of-00047.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..83b34ccf07b6134d0d1384c69d76898a9331451b
--- /dev/null
+++ b/model-00040-of-00047.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:103a33ce23be80b19b18c51eecb1cf15cd49584e22b9138b1d358e8720b2abf3
+size 9995591008
diff --git a/model-00041-of-00047.safetensors b/model-00041-of-00047.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..164a106b7ec9553170228af724c695dbe6d0780c
--- /dev/null
+++ b/model-00041-of-00047.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4d398e69fdd5ca91aa580ded9fbfd1032321e3bdc08b32f647cd27b16ebc0986
+size 9998081416
diff --git a/model-00042-of-00047.safetensors b/model-00042-of-00047.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..ec3f55ed55516f9b4b51f90d68f8d40a71a44ec1
--- /dev/null
+++ b/model-00042-of-00047.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:573afa976e2f6c63a4b7256d53d5a7f5d1070b2dc144e2aec3d96b44256b9dec
+size 9881695296
diff --git a/model-00043-of-00047.safetensors b/model-00043-of-00047.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..3fa0c69c1d1d2b71a6d6c584a02a8136ba56c109
--- /dev/null
+++ b/model-00043-of-00047.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a459ccf072fc918b219766eb167d773203cdd92694d8d40ebf44acb2e4de5814
+size 10000552944
diff --git a/model-00044-of-00047.safetensors b/model-00044-of-00047.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..38e7889c045eef5824bdc1bbc452a58dba4ff5a4
--- /dev/null
+++ b/model-00044-of-00047.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5ef1a45b4d69bd8a3d2e4d4e895242b920ad3248e2869276f3f06bf43923d18e
+size 9998081424
diff --git a/model-00045-of-00047.safetensors b/model-00045-of-00047.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..d76152257fe7556f1ffca425a7ae06b1aefbf847
--- /dev/null
+++ b/model-00045-of-00047.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0fa60cd2ffddfcaf82e9dcc36fc42e1ec0db623966154b84b7f041c2ab8688e1
+size 9980927784
diff --git a/model-00046-of-00047.safetensors b/model-00046-of-00047.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..84e338f459a5b87eae34359e8ff1b9da363e838a
--- /dev/null
+++ b/model-00046-of-00047.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6edd678c7c8998860d75c8a7abb431084800a8d11e6d8ad6b88fe5215b5c309b
+size 9990882136
diff --git a/model-00047-of-00047.safetensors b/model-00047-of-00047.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..aad6c4172063cb04ed1c867e6aa81c9e76b1769e
--- /dev/null
+++ b/model-00047-of-00047.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ca2dfdb8f16f56c425de8240409bd96f1bb3ee9493505acf2d9b6f42b833c1ba
+size 5108325840
diff --git a/model.safetensors.index.json b/model.safetensors.index.json
new file mode 100644
index 0000000000000000000000000000000000000000..0b72f5ba0e99a9b9d31d53191b2eea68dce4e84b
--- /dev/null
+++ b/model.safetensors.index.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2aa8397b501d9f6a232d153f328feb912f813c389061aac4cf72b04914fa5b74
+size 22194492
diff --git a/tokenizer.json b/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..aba40197a4cdb5607f4ab7a05fb0a4ee8054fd6d
--- /dev/null
+++ b/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:19e773648cb4e65de8660ea6365e10acca112d42a854923df93db4a6f333a82d
+size 20217442
diff --git a/tokenizer_config.json b/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..0891a18eef6fc3d26dd2e87796ed39c7f91e3a77
--- /dev/null
+++ b/tokenizer_config.json
@@ -0,0 +1,34 @@
+{
+  "backend": "tokenizers",
+  "clean_up_tokenization_spaces": false,
+  "do_lower_case": false,
+  "eos_token": "<|endoftext|>",
+  "extra_special_tokens": [
+    "<|endoftext|>",
+    "[MASK]",
+    "[gMASK]",
+    "[sMASK]",
+    "<sop>",
+    "<eop>",
+    "<|system|>",
+    "<|user|>",
+    "<|assistant|>",
+    "<|observation|>",
+    "<|begin_of_image|>",
+    "<|end_of_image|>",
+    "<|begin_of_video|>",
+    "<|end_of_video|>",
+    "<|begin_of_audio|>",
+    "<|end_of_audio|>",
+    "<|begin_of_transcription|>",
+    "<|end_of_transcription|>"
+  ],
+  "is_local": true,
+  "local_files_only": false,
+  "model_max_length": 1048576,
+  "model_specific_special_tokens": {},
+  "pad_token": "<|endoftext|>",
+  "padding_side": "left",
+  "remove_space": false,
+  "tokenizer_class": "TokenizersBackend"
+}