diff --git "a/modelopt-v3/_quant_summary/.quant_summary.txt" "b/modelopt-v3/_quant_summary/.quant_summary.txt" new file mode 100644--- /dev/null +++ "b/modelopt-v3/_quant_summary/.quant_summary.txt" @@ -0,0 +1,1363 @@ +model.visual.patch_embed.proj.input_quantizer TensorQuantizer(disabled) +model.visual.patch_embed.proj.output_quantizer TensorQuantizer(disabled) +model.visual.patch_embed.proj.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.0.attn.qkv.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.0.attn.qkv.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.0.attn.qkv.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.0.attn.proj.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.0.attn.proj.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.0.attn.proj.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.0.attn.q_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.0.attn.k_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.0.attn.v_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.0.attn.softmax_quantizer TensorQuantizer(disabled) +model.visual.blocks.0.mlp.linear_fc1.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.0.mlp.linear_fc1.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.0.mlp.linear_fc1.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.0.mlp.linear_fc2.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.0.mlp.linear_fc2.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.0.mlp.linear_fc2.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.1.attn.qkv.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.1.attn.qkv.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.1.attn.qkv.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.1.attn.proj.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.1.attn.proj.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.1.attn.proj.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.1.attn.q_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.1.attn.k_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.1.attn.v_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.1.attn.softmax_quantizer TensorQuantizer(disabled) +model.visual.blocks.1.mlp.linear_fc1.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.1.mlp.linear_fc1.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.1.mlp.linear_fc1.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.1.mlp.linear_fc2.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.1.mlp.linear_fc2.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.1.mlp.linear_fc2.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.2.attn.qkv.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.2.attn.qkv.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.2.attn.qkv.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.2.attn.proj.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.2.attn.proj.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.2.attn.proj.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.2.attn.q_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.2.attn.k_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.2.attn.v_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.2.attn.softmax_quantizer TensorQuantizer(disabled) +model.visual.blocks.2.mlp.linear_fc1.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.2.mlp.linear_fc1.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.2.mlp.linear_fc1.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.2.mlp.linear_fc2.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.2.mlp.linear_fc2.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.2.mlp.linear_fc2.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.3.attn.qkv.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.3.attn.qkv.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.3.attn.qkv.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.3.attn.proj.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.3.attn.proj.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.3.attn.proj.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.3.attn.q_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.3.attn.k_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.3.attn.v_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.3.attn.softmax_quantizer TensorQuantizer(disabled) +model.visual.blocks.3.mlp.linear_fc1.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.3.mlp.linear_fc1.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.3.mlp.linear_fc1.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.3.mlp.linear_fc2.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.3.mlp.linear_fc2.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.3.mlp.linear_fc2.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.4.attn.qkv.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.4.attn.qkv.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.4.attn.qkv.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.4.attn.proj.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.4.attn.proj.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.4.attn.proj.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.4.attn.q_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.4.attn.k_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.4.attn.v_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.4.attn.softmax_quantizer TensorQuantizer(disabled) +model.visual.blocks.4.mlp.linear_fc1.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.4.mlp.linear_fc1.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.4.mlp.linear_fc1.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.4.mlp.linear_fc2.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.4.mlp.linear_fc2.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.4.mlp.linear_fc2.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.5.attn.qkv.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.5.attn.qkv.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.5.attn.qkv.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.5.attn.proj.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.5.attn.proj.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.5.attn.proj.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.5.attn.q_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.5.attn.k_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.5.attn.v_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.5.attn.softmax_quantizer TensorQuantizer(disabled) +model.visual.blocks.5.mlp.linear_fc1.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.5.mlp.linear_fc1.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.5.mlp.linear_fc1.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.5.mlp.linear_fc2.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.5.mlp.linear_fc2.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.5.mlp.linear_fc2.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.6.attn.qkv.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.6.attn.qkv.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.6.attn.qkv.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.6.attn.proj.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.6.attn.proj.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.6.attn.proj.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.6.attn.q_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.6.attn.k_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.6.attn.v_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.6.attn.softmax_quantizer TensorQuantizer(disabled) +model.visual.blocks.6.mlp.linear_fc1.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.6.mlp.linear_fc1.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.6.mlp.linear_fc1.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.6.mlp.linear_fc2.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.6.mlp.linear_fc2.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.6.mlp.linear_fc2.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.7.attn.qkv.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.7.attn.qkv.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.7.attn.qkv.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.7.attn.proj.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.7.attn.proj.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.7.attn.proj.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.7.attn.q_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.7.attn.k_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.7.attn.v_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.7.attn.softmax_quantizer TensorQuantizer(disabled) +model.visual.blocks.7.mlp.linear_fc1.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.7.mlp.linear_fc1.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.7.mlp.linear_fc1.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.7.mlp.linear_fc2.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.7.mlp.linear_fc2.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.7.mlp.linear_fc2.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.8.attn.qkv.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.8.attn.qkv.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.8.attn.qkv.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.8.attn.proj.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.8.attn.proj.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.8.attn.proj.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.8.attn.q_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.8.attn.k_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.8.attn.v_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.8.attn.softmax_quantizer TensorQuantizer(disabled) +model.visual.blocks.8.mlp.linear_fc1.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.8.mlp.linear_fc1.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.8.mlp.linear_fc1.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.8.mlp.linear_fc2.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.8.mlp.linear_fc2.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.8.mlp.linear_fc2.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.9.attn.qkv.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.9.attn.qkv.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.9.attn.qkv.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.9.attn.proj.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.9.attn.proj.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.9.attn.proj.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.9.attn.q_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.9.attn.k_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.9.attn.v_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.9.attn.softmax_quantizer TensorQuantizer(disabled) +model.visual.blocks.9.mlp.linear_fc1.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.9.mlp.linear_fc1.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.9.mlp.linear_fc1.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.9.mlp.linear_fc2.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.9.mlp.linear_fc2.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.9.mlp.linear_fc2.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.10.attn.qkv.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.10.attn.qkv.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.10.attn.qkv.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.10.attn.proj.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.10.attn.proj.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.10.attn.proj.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.10.attn.q_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.10.attn.k_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.10.attn.v_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.10.attn.softmax_quantizer TensorQuantizer(disabled) +model.visual.blocks.10.mlp.linear_fc1.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.10.mlp.linear_fc1.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.10.mlp.linear_fc1.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.10.mlp.linear_fc2.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.10.mlp.linear_fc2.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.10.mlp.linear_fc2.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.11.attn.qkv.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.11.attn.qkv.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.11.attn.qkv.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.11.attn.proj.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.11.attn.proj.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.11.attn.proj.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.11.attn.q_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.11.attn.k_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.11.attn.v_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.11.attn.softmax_quantizer TensorQuantizer(disabled) +model.visual.blocks.11.mlp.linear_fc1.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.11.mlp.linear_fc1.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.11.mlp.linear_fc1.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.11.mlp.linear_fc2.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.11.mlp.linear_fc2.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.11.mlp.linear_fc2.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.12.attn.qkv.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.12.attn.qkv.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.12.attn.qkv.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.12.attn.proj.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.12.attn.proj.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.12.attn.proj.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.12.attn.q_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.12.attn.k_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.12.attn.v_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.12.attn.softmax_quantizer TensorQuantizer(disabled) +model.visual.blocks.12.mlp.linear_fc1.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.12.mlp.linear_fc1.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.12.mlp.linear_fc1.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.12.mlp.linear_fc2.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.12.mlp.linear_fc2.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.12.mlp.linear_fc2.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.13.attn.qkv.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.13.attn.qkv.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.13.attn.qkv.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.13.attn.proj.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.13.attn.proj.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.13.attn.proj.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.13.attn.q_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.13.attn.k_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.13.attn.v_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.13.attn.softmax_quantizer TensorQuantizer(disabled) +model.visual.blocks.13.mlp.linear_fc1.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.13.mlp.linear_fc1.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.13.mlp.linear_fc1.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.13.mlp.linear_fc2.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.13.mlp.linear_fc2.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.13.mlp.linear_fc2.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.14.attn.qkv.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.14.attn.qkv.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.14.attn.qkv.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.14.attn.proj.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.14.attn.proj.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.14.attn.proj.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.14.attn.q_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.14.attn.k_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.14.attn.v_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.14.attn.softmax_quantizer TensorQuantizer(disabled) +model.visual.blocks.14.mlp.linear_fc1.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.14.mlp.linear_fc1.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.14.mlp.linear_fc1.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.14.mlp.linear_fc2.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.14.mlp.linear_fc2.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.14.mlp.linear_fc2.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.15.attn.qkv.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.15.attn.qkv.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.15.attn.qkv.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.15.attn.proj.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.15.attn.proj.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.15.attn.proj.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.15.attn.q_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.15.attn.k_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.15.attn.v_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.15.attn.softmax_quantizer TensorQuantizer(disabled) +model.visual.blocks.15.mlp.linear_fc1.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.15.mlp.linear_fc1.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.15.mlp.linear_fc1.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.15.mlp.linear_fc2.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.15.mlp.linear_fc2.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.15.mlp.linear_fc2.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.16.attn.qkv.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.16.attn.qkv.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.16.attn.qkv.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.16.attn.proj.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.16.attn.proj.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.16.attn.proj.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.16.attn.q_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.16.attn.k_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.16.attn.v_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.16.attn.softmax_quantizer TensorQuantizer(disabled) +model.visual.blocks.16.mlp.linear_fc1.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.16.mlp.linear_fc1.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.16.mlp.linear_fc1.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.16.mlp.linear_fc2.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.16.mlp.linear_fc2.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.16.mlp.linear_fc2.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.17.attn.qkv.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.17.attn.qkv.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.17.attn.qkv.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.17.attn.proj.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.17.attn.proj.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.17.attn.proj.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.17.attn.q_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.17.attn.k_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.17.attn.v_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.17.attn.softmax_quantizer TensorQuantizer(disabled) +model.visual.blocks.17.mlp.linear_fc1.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.17.mlp.linear_fc1.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.17.mlp.linear_fc1.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.17.mlp.linear_fc2.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.17.mlp.linear_fc2.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.17.mlp.linear_fc2.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.18.attn.qkv.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.18.attn.qkv.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.18.attn.qkv.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.18.attn.proj.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.18.attn.proj.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.18.attn.proj.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.18.attn.q_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.18.attn.k_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.18.attn.v_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.18.attn.softmax_quantizer TensorQuantizer(disabled) +model.visual.blocks.18.mlp.linear_fc1.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.18.mlp.linear_fc1.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.18.mlp.linear_fc1.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.18.mlp.linear_fc2.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.18.mlp.linear_fc2.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.18.mlp.linear_fc2.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.19.attn.qkv.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.19.attn.qkv.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.19.attn.qkv.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.19.attn.proj.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.19.attn.proj.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.19.attn.proj.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.19.attn.q_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.19.attn.k_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.19.attn.v_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.19.attn.softmax_quantizer TensorQuantizer(disabled) +model.visual.blocks.19.mlp.linear_fc1.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.19.mlp.linear_fc1.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.19.mlp.linear_fc1.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.19.mlp.linear_fc2.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.19.mlp.linear_fc2.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.19.mlp.linear_fc2.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.20.attn.qkv.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.20.attn.qkv.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.20.attn.qkv.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.20.attn.proj.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.20.attn.proj.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.20.attn.proj.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.20.attn.q_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.20.attn.k_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.20.attn.v_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.20.attn.softmax_quantizer TensorQuantizer(disabled) +model.visual.blocks.20.mlp.linear_fc1.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.20.mlp.linear_fc1.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.20.mlp.linear_fc1.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.20.mlp.linear_fc2.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.20.mlp.linear_fc2.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.20.mlp.linear_fc2.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.21.attn.qkv.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.21.attn.qkv.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.21.attn.qkv.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.21.attn.proj.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.21.attn.proj.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.21.attn.proj.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.21.attn.q_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.21.attn.k_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.21.attn.v_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.21.attn.softmax_quantizer TensorQuantizer(disabled) +model.visual.blocks.21.mlp.linear_fc1.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.21.mlp.linear_fc1.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.21.mlp.linear_fc1.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.21.mlp.linear_fc2.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.21.mlp.linear_fc2.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.21.mlp.linear_fc2.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.22.attn.qkv.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.22.attn.qkv.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.22.attn.qkv.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.22.attn.proj.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.22.attn.proj.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.22.attn.proj.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.22.attn.q_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.22.attn.k_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.22.attn.v_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.22.attn.softmax_quantizer TensorQuantizer(disabled) +model.visual.blocks.22.mlp.linear_fc1.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.22.mlp.linear_fc1.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.22.mlp.linear_fc1.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.22.mlp.linear_fc2.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.22.mlp.linear_fc2.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.22.mlp.linear_fc2.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.23.attn.qkv.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.23.attn.qkv.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.23.attn.qkv.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.23.attn.proj.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.23.attn.proj.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.23.attn.proj.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.23.attn.q_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.23.attn.k_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.23.attn.v_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.23.attn.softmax_quantizer TensorQuantizer(disabled) +model.visual.blocks.23.mlp.linear_fc1.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.23.mlp.linear_fc1.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.23.mlp.linear_fc1.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.23.mlp.linear_fc2.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.23.mlp.linear_fc2.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.23.mlp.linear_fc2.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.24.attn.qkv.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.24.attn.qkv.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.24.attn.qkv.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.24.attn.proj.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.24.attn.proj.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.24.attn.proj.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.24.attn.q_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.24.attn.k_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.24.attn.v_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.24.attn.softmax_quantizer TensorQuantizer(disabled) +model.visual.blocks.24.mlp.linear_fc1.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.24.mlp.linear_fc1.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.24.mlp.linear_fc1.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.24.mlp.linear_fc2.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.24.mlp.linear_fc2.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.24.mlp.linear_fc2.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.25.attn.qkv.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.25.attn.qkv.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.25.attn.qkv.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.25.attn.proj.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.25.attn.proj.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.25.attn.proj.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.25.attn.q_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.25.attn.k_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.25.attn.v_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.25.attn.softmax_quantizer TensorQuantizer(disabled) +model.visual.blocks.25.mlp.linear_fc1.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.25.mlp.linear_fc1.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.25.mlp.linear_fc1.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.25.mlp.linear_fc2.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.25.mlp.linear_fc2.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.25.mlp.linear_fc2.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.26.attn.qkv.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.26.attn.qkv.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.26.attn.qkv.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.26.attn.proj.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.26.attn.proj.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.26.attn.proj.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.26.attn.q_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.26.attn.k_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.26.attn.v_bmm_quantizer TensorQuantizer(disabled) +model.visual.blocks.26.attn.softmax_quantizer TensorQuantizer(disabled) +model.visual.blocks.26.mlp.linear_fc1.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.26.mlp.linear_fc1.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.26.mlp.linear_fc1.weight_quantizer TensorQuantizer(disabled) +model.visual.blocks.26.mlp.linear_fc2.input_quantizer TensorQuantizer(disabled) +model.visual.blocks.26.mlp.linear_fc2.output_quantizer TensorQuantizer(disabled) +model.visual.blocks.26.mlp.linear_fc2.weight_quantizer TensorQuantizer(disabled) +model.visual.merger.linear_fc1.input_quantizer TensorQuantizer(disabled) +model.visual.merger.linear_fc1.output_quantizer TensorQuantizer(disabled) +model.visual.merger.linear_fc1.weight_quantizer TensorQuantizer(disabled) +model.visual.merger.linear_fc2.input_quantizer TensorQuantizer(disabled) +model.visual.merger.linear_fc2.output_quantizer TensorQuantizer(disabled) +model.visual.merger.linear_fc2.weight_quantizer TensorQuantizer(disabled) +model.visual.deepstack_merger_list.0.linear_fc1.input_quantizer TensorQuantizer(disabled) +model.visual.deepstack_merger_list.0.linear_fc1.output_quantizer TensorQuantizer(disabled) +model.visual.deepstack_merger_list.0.linear_fc1.weight_quantizer TensorQuantizer(disabled) +model.visual.deepstack_merger_list.0.linear_fc2.input_quantizer TensorQuantizer(disabled) +model.visual.deepstack_merger_list.0.linear_fc2.output_quantizer TensorQuantizer(disabled) +model.visual.deepstack_merger_list.0.linear_fc2.weight_quantizer TensorQuantizer(disabled) +model.visual.deepstack_merger_list.1.linear_fc1.input_quantizer TensorQuantizer(disabled) +model.visual.deepstack_merger_list.1.linear_fc1.output_quantizer TensorQuantizer(disabled) +model.visual.deepstack_merger_list.1.linear_fc1.weight_quantizer TensorQuantizer(disabled) +model.visual.deepstack_merger_list.1.linear_fc2.input_quantizer TensorQuantizer(disabled) +model.visual.deepstack_merger_list.1.linear_fc2.output_quantizer TensorQuantizer(disabled) +model.visual.deepstack_merger_list.1.linear_fc2.weight_quantizer TensorQuantizer(disabled) +model.visual.deepstack_merger_list.2.linear_fc1.input_quantizer TensorQuantizer(disabled) +model.visual.deepstack_merger_list.2.linear_fc1.output_quantizer TensorQuantizer(disabled) +model.visual.deepstack_merger_list.2.linear_fc1.weight_quantizer TensorQuantizer(disabled) +model.visual.deepstack_merger_list.2.linear_fc2.input_quantizer TensorQuantizer(disabled) +model.visual.deepstack_merger_list.2.linear_fc2.output_quantizer TensorQuantizer(disabled) +model.visual.deepstack_merger_list.2.linear_fc2.weight_quantizer TensorQuantizer(disabled) +model.language_model.layers.0.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.0.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.0.self_attn.q_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.0.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.0.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.0.self_attn.k_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.0.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.0.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.0.self_attn.v_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.0.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.0.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.0.self_attn.o_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.0.self_attn.q_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.0.self_attn.k_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.0.self_attn.v_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.0.self_attn.softmax_quantizer TensorQuantizer(disabled) +model.language_model.layers.0.mlp.gate_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.0.mlp.gate_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.0.mlp.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.0.mlp.up_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.0.mlp.up_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.0.mlp.up_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.0.mlp.down_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.0.mlp.down_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.0.mlp.down_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.1.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.1.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.1.self_attn.q_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.1.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.1.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.1.self_attn.k_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.1.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.1.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.1.self_attn.v_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.1.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.1.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.1.self_attn.o_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.1.self_attn.q_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.1.self_attn.k_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.1.self_attn.v_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.1.self_attn.softmax_quantizer TensorQuantizer(disabled) +model.language_model.layers.1.mlp.gate_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.1.mlp.gate_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.1.mlp.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.1.mlp.up_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.1.mlp.up_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.1.mlp.up_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.1.mlp.down_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.1.mlp.down_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.1.mlp.down_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.2.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.2.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.2.self_attn.q_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.2.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.2.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.2.self_attn.k_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.2.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.2.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.2.self_attn.v_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.2.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.2.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.2.self_attn.o_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.2.self_attn.q_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.2.self_attn.k_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.2.self_attn.v_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.2.self_attn.softmax_quantizer TensorQuantizer(disabled) +model.language_model.layers.2.mlp.gate_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.2.mlp.gate_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.2.mlp.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.2.mlp.up_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.2.mlp.up_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.2.mlp.up_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.2.mlp.down_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.2.mlp.down_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.2.mlp.down_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.3.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.3.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.3.self_attn.q_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.3.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.3.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.3.self_attn.k_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.3.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.3.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.3.self_attn.v_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.3.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.3.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.3.self_attn.o_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.3.self_attn.q_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.3.self_attn.k_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.3.self_attn.v_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.3.self_attn.softmax_quantizer TensorQuantizer(disabled) +model.language_model.layers.3.mlp.gate_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.3.mlp.gate_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.3.mlp.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.3.mlp.up_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.3.mlp.up_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.3.mlp.up_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.3.mlp.down_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.3.mlp.down_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.3.mlp.down_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.4.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.4.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.4.self_attn.q_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.4.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.4.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.4.self_attn.k_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.4.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.4.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.4.self_attn.v_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.4.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.4.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.4.self_attn.o_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.4.self_attn.q_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.4.self_attn.k_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.4.self_attn.v_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.4.self_attn.softmax_quantizer TensorQuantizer(disabled) +model.language_model.layers.4.mlp.gate_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.4.mlp.gate_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.4.mlp.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.4.mlp.up_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.4.mlp.up_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.4.mlp.up_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.4.mlp.down_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.4.mlp.down_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.4.mlp.down_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.5.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.5.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.5.self_attn.q_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.5.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.5.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.5.self_attn.k_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.5.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.5.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.5.self_attn.v_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.5.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.5.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.5.self_attn.o_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.5.self_attn.q_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.5.self_attn.k_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.5.self_attn.v_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.5.self_attn.softmax_quantizer TensorQuantizer(disabled) +model.language_model.layers.5.mlp.gate_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.5.mlp.gate_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.5.mlp.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.5.mlp.up_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.5.mlp.up_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.5.mlp.up_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.5.mlp.down_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.5.mlp.down_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.5.mlp.down_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.6.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.6.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.6.self_attn.q_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.6.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.6.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.6.self_attn.k_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.6.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.6.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.6.self_attn.v_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.6.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.6.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.6.self_attn.o_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.6.self_attn.q_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.6.self_attn.k_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.6.self_attn.v_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.6.self_attn.softmax_quantizer TensorQuantizer(disabled) +model.language_model.layers.6.mlp.gate_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.6.mlp.gate_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.6.mlp.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.6.mlp.up_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.6.mlp.up_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.6.mlp.up_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.6.mlp.down_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.6.mlp.down_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.6.mlp.down_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.7.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.7.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.7.self_attn.q_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.7.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.7.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.7.self_attn.k_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.7.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.7.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.7.self_attn.v_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.7.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.7.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.7.self_attn.o_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.7.self_attn.q_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.7.self_attn.k_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.7.self_attn.v_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.7.self_attn.softmax_quantizer TensorQuantizer(disabled) +model.language_model.layers.7.mlp.gate_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.7.mlp.gate_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.7.mlp.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.7.mlp.up_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.7.mlp.up_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.7.mlp.up_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.7.mlp.down_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.7.mlp.down_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.7.mlp.down_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.8.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.8.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.8.self_attn.q_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.8.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.8.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.8.self_attn.k_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.8.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.8.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.8.self_attn.v_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.8.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.8.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.8.self_attn.o_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.8.self_attn.q_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.8.self_attn.k_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.8.self_attn.v_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.8.self_attn.softmax_quantizer TensorQuantizer(disabled) +model.language_model.layers.8.mlp.gate_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.8.mlp.gate_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.8.mlp.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.8.mlp.up_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.8.mlp.up_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.8.mlp.up_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.8.mlp.down_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.8.mlp.down_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.8.mlp.down_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.9.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.9.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.9.self_attn.q_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.9.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.9.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.9.self_attn.k_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.9.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.9.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.9.self_attn.v_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.9.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.9.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.9.self_attn.o_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.9.self_attn.q_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.9.self_attn.k_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.9.self_attn.v_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.9.self_attn.softmax_quantizer TensorQuantizer(disabled) +model.language_model.layers.9.mlp.gate_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.9.mlp.gate_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.9.mlp.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.9.mlp.up_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.9.mlp.up_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.9.mlp.up_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.9.mlp.down_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.9.mlp.down_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.9.mlp.down_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.10.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.10.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.10.self_attn.q_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.10.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.10.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.10.self_attn.k_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.10.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.10.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.10.self_attn.v_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.10.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.10.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.10.self_attn.o_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.10.self_attn.q_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.10.self_attn.k_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.10.self_attn.v_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.10.self_attn.softmax_quantizer TensorQuantizer(disabled) +model.language_model.layers.10.mlp.gate_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.10.mlp.gate_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.10.mlp.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.10.mlp.up_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.10.mlp.up_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.10.mlp.up_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.10.mlp.down_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.10.mlp.down_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.10.mlp.down_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.11.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.11.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.11.self_attn.q_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.11.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.11.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.11.self_attn.k_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.11.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.11.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.11.self_attn.v_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.11.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.11.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.11.self_attn.o_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.11.self_attn.q_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.11.self_attn.k_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.11.self_attn.v_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.11.self_attn.softmax_quantizer TensorQuantizer(disabled) +model.language_model.layers.11.mlp.gate_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.11.mlp.gate_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.11.mlp.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.11.mlp.up_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.11.mlp.up_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.11.mlp.up_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.11.mlp.down_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.11.mlp.down_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.11.mlp.down_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.12.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.12.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.12.self_attn.q_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.12.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.12.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.12.self_attn.k_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.12.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.12.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.12.self_attn.v_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.12.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.12.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.12.self_attn.o_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.12.self_attn.q_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.12.self_attn.k_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.12.self_attn.v_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.12.self_attn.softmax_quantizer TensorQuantizer(disabled) +model.language_model.layers.12.mlp.gate_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.12.mlp.gate_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.12.mlp.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.12.mlp.up_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.12.mlp.up_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.12.mlp.up_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.12.mlp.down_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.12.mlp.down_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.12.mlp.down_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.13.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.13.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.13.self_attn.q_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.13.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.13.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.13.self_attn.k_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.13.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.13.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.13.self_attn.v_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.13.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.13.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.13.self_attn.o_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.13.self_attn.q_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.13.self_attn.k_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.13.self_attn.v_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.13.self_attn.softmax_quantizer TensorQuantizer(disabled) +model.language_model.layers.13.mlp.gate_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.13.mlp.gate_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.13.mlp.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.13.mlp.up_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.13.mlp.up_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.13.mlp.up_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.13.mlp.down_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.13.mlp.down_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.13.mlp.down_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.14.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.14.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.14.self_attn.q_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.14.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.14.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.14.self_attn.k_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.14.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.14.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.14.self_attn.v_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.14.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.14.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.14.self_attn.o_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.14.self_attn.q_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.14.self_attn.k_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.14.self_attn.v_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.14.self_attn.softmax_quantizer TensorQuantizer(disabled) +model.language_model.layers.14.mlp.gate_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.14.mlp.gate_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.14.mlp.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.14.mlp.up_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.14.mlp.up_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.14.mlp.up_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.14.mlp.down_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.14.mlp.down_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.14.mlp.down_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.15.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.15.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.15.self_attn.q_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.15.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.15.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.15.self_attn.k_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.15.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.15.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.15.self_attn.v_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.15.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.15.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.15.self_attn.o_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.15.self_attn.q_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.15.self_attn.k_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.15.self_attn.v_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.15.self_attn.softmax_quantizer TensorQuantizer(disabled) +model.language_model.layers.15.mlp.gate_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.15.mlp.gate_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.15.mlp.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.15.mlp.up_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.15.mlp.up_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.15.mlp.up_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.15.mlp.down_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.15.mlp.down_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.15.mlp.down_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.16.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.16.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.16.self_attn.q_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.16.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.16.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.16.self_attn.k_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.16.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.16.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.16.self_attn.v_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.16.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.16.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.16.self_attn.o_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.16.self_attn.q_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.16.self_attn.k_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.16.self_attn.v_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.16.self_attn.softmax_quantizer TensorQuantizer(disabled) +model.language_model.layers.16.mlp.gate_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.16.mlp.gate_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.16.mlp.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.16.mlp.up_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.16.mlp.up_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.16.mlp.up_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.16.mlp.down_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.16.mlp.down_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.16.mlp.down_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.17.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.17.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.17.self_attn.q_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.17.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.17.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.17.self_attn.k_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.17.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.17.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.17.self_attn.v_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.17.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.17.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.17.self_attn.o_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.17.self_attn.q_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.17.self_attn.k_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.17.self_attn.v_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.17.self_attn.softmax_quantizer TensorQuantizer(disabled) +model.language_model.layers.17.mlp.gate_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.17.mlp.gate_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.17.mlp.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.17.mlp.up_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.17.mlp.up_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.17.mlp.up_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.17.mlp.down_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.17.mlp.down_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.17.mlp.down_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.18.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.18.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.18.self_attn.q_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.18.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.18.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.18.self_attn.k_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.18.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.18.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.18.self_attn.v_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.18.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.18.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.18.self_attn.o_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.18.self_attn.q_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.18.self_attn.k_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.18.self_attn.v_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.18.self_attn.softmax_quantizer TensorQuantizer(disabled) +model.language_model.layers.18.mlp.gate_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.18.mlp.gate_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.18.mlp.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.18.mlp.up_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.18.mlp.up_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.18.mlp.up_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.18.mlp.down_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.18.mlp.down_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.18.mlp.down_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.19.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.19.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.19.self_attn.q_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.19.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.19.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.19.self_attn.k_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.19.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.19.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.19.self_attn.v_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.19.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.19.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.19.self_attn.o_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.19.self_attn.q_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.19.self_attn.k_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.19.self_attn.v_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.19.self_attn.softmax_quantizer TensorQuantizer(disabled) +model.language_model.layers.19.mlp.gate_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.19.mlp.gate_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.19.mlp.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.19.mlp.up_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.19.mlp.up_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.19.mlp.up_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.19.mlp.down_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.19.mlp.down_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.19.mlp.down_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.20.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.20.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.20.self_attn.q_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.20.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.20.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.20.self_attn.k_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.20.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.20.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.20.self_attn.v_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.20.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.20.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.20.self_attn.o_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.20.self_attn.q_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.20.self_attn.k_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.20.self_attn.v_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.20.self_attn.softmax_quantizer TensorQuantizer(disabled) +model.language_model.layers.20.mlp.gate_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.20.mlp.gate_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.20.mlp.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.20.mlp.up_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.20.mlp.up_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.20.mlp.up_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.20.mlp.down_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.20.mlp.down_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.20.mlp.down_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.21.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.21.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.21.self_attn.q_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.21.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.21.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.21.self_attn.k_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.21.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.21.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.21.self_attn.v_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.21.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.21.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.21.self_attn.o_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.21.self_attn.q_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.21.self_attn.k_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.21.self_attn.v_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.21.self_attn.softmax_quantizer TensorQuantizer(disabled) +model.language_model.layers.21.mlp.gate_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.21.mlp.gate_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.21.mlp.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.21.mlp.up_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.21.mlp.up_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.21.mlp.up_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.21.mlp.down_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.21.mlp.down_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.21.mlp.down_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.22.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.22.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.22.self_attn.q_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.22.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.22.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.22.self_attn.k_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.22.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.22.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.22.self_attn.v_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.22.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.22.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.22.self_attn.o_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.22.self_attn.q_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.22.self_attn.k_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.22.self_attn.v_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.22.self_attn.softmax_quantizer TensorQuantizer(disabled) +model.language_model.layers.22.mlp.gate_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.22.mlp.gate_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.22.mlp.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.22.mlp.up_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.22.mlp.up_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.22.mlp.up_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.22.mlp.down_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.22.mlp.down_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.22.mlp.down_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.23.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.23.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.23.self_attn.q_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.23.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.23.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.23.self_attn.k_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.23.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.23.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.23.self_attn.v_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.23.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.23.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.23.self_attn.o_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.23.self_attn.q_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.23.self_attn.k_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.23.self_attn.v_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.23.self_attn.softmax_quantizer TensorQuantizer(disabled) +model.language_model.layers.23.mlp.gate_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.23.mlp.gate_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.23.mlp.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.23.mlp.up_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.23.mlp.up_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.23.mlp.up_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.23.mlp.down_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.23.mlp.down_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.23.mlp.down_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.24.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.24.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.24.self_attn.q_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.24.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.24.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.24.self_attn.k_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.24.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.24.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.24.self_attn.v_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.24.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.24.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.24.self_attn.o_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.24.self_attn.q_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.24.self_attn.k_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.24.self_attn.v_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.24.self_attn.softmax_quantizer TensorQuantizer(disabled) +model.language_model.layers.24.mlp.gate_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.24.mlp.gate_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.24.mlp.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.24.mlp.up_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.24.mlp.up_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.24.mlp.up_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.24.mlp.down_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.24.mlp.down_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.24.mlp.down_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.25.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.25.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.25.self_attn.q_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.25.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.25.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.25.self_attn.k_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.25.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.25.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.25.self_attn.v_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.25.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.25.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.25.self_attn.o_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.25.self_attn.q_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.25.self_attn.k_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.25.self_attn.v_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.25.self_attn.softmax_quantizer TensorQuantizer(disabled) +model.language_model.layers.25.mlp.gate_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.25.mlp.gate_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.25.mlp.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.25.mlp.up_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.25.mlp.up_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.25.mlp.up_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.25.mlp.down_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.25.mlp.down_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.25.mlp.down_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.26.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.26.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.26.self_attn.q_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.26.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.26.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.26.self_attn.k_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.26.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.26.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.26.self_attn.v_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.26.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.26.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.26.self_attn.o_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.26.self_attn.q_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.26.self_attn.k_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.26.self_attn.v_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.26.self_attn.softmax_quantizer TensorQuantizer(disabled) +model.language_model.layers.26.mlp.gate_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.26.mlp.gate_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.26.mlp.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.26.mlp.up_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.26.mlp.up_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.26.mlp.up_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.26.mlp.down_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.26.mlp.down_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.26.mlp.down_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.27.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.27.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.27.self_attn.q_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.27.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.27.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.27.self_attn.k_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.27.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.27.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.27.self_attn.v_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.27.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.27.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.27.self_attn.o_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.27.self_attn.q_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.27.self_attn.k_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.27.self_attn.v_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.27.self_attn.softmax_quantizer TensorQuantizer(disabled) +model.language_model.layers.27.mlp.gate_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.27.mlp.gate_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.27.mlp.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.27.mlp.up_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.27.mlp.up_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.27.mlp.up_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.27.mlp.down_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.27.mlp.down_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.27.mlp.down_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.28.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.28.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.28.self_attn.q_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.28.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.28.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.28.self_attn.k_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.28.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.28.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.28.self_attn.v_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.28.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.28.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.28.self_attn.o_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.28.self_attn.q_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.28.self_attn.k_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.28.self_attn.v_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.28.self_attn.softmax_quantizer TensorQuantizer(disabled) +model.language_model.layers.28.mlp.gate_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.28.mlp.gate_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.28.mlp.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.28.mlp.up_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.28.mlp.up_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.28.mlp.up_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.28.mlp.down_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.28.mlp.down_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.28.mlp.down_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.29.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.29.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.29.self_attn.q_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.29.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.29.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.29.self_attn.k_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.29.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.29.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.29.self_attn.v_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.29.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.29.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.29.self_attn.o_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.29.self_attn.q_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.29.self_attn.k_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.29.self_attn.v_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.29.self_attn.softmax_quantizer TensorQuantizer(disabled) +model.language_model.layers.29.mlp.gate_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.29.mlp.gate_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.29.mlp.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.29.mlp.up_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.29.mlp.up_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.29.mlp.up_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.29.mlp.down_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.29.mlp.down_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.29.mlp.down_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.30.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.30.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.30.self_attn.q_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.30.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.30.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.30.self_attn.k_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.30.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.30.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.30.self_attn.v_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.30.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.30.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.30.self_attn.o_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.30.self_attn.q_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.30.self_attn.k_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.30.self_attn.v_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.30.self_attn.softmax_quantizer TensorQuantizer(disabled) +model.language_model.layers.30.mlp.gate_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.30.mlp.gate_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.30.mlp.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.30.mlp.up_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.30.mlp.up_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.30.mlp.up_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.30.mlp.down_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.30.mlp.down_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.30.mlp.down_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.31.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.31.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.31.self_attn.q_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.31.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.31.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.31.self_attn.k_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.31.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.31.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.31.self_attn.v_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.31.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.31.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.31.self_attn.o_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.31.self_attn.q_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.31.self_attn.k_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.31.self_attn.v_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.31.self_attn.softmax_quantizer TensorQuantizer(disabled) +model.language_model.layers.31.mlp.gate_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.31.mlp.gate_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.31.mlp.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.31.mlp.up_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.31.mlp.up_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.31.mlp.up_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.31.mlp.down_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.31.mlp.down_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.31.mlp.down_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.32.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.32.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.32.self_attn.q_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.32.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.32.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.32.self_attn.k_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.32.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.32.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.32.self_attn.v_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.32.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.32.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.32.self_attn.o_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.32.self_attn.q_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.32.self_attn.k_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.32.self_attn.v_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.32.self_attn.softmax_quantizer TensorQuantizer(disabled) +model.language_model.layers.32.mlp.gate_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.32.mlp.gate_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.32.mlp.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.32.mlp.up_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.32.mlp.up_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.32.mlp.up_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.32.mlp.down_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.32.mlp.down_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.32.mlp.down_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.33.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.33.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.33.self_attn.q_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.33.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.33.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.33.self_attn.k_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.33.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.33.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.33.self_attn.v_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.33.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.33.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.33.self_attn.o_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.33.self_attn.q_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.33.self_attn.k_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.33.self_attn.v_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.33.self_attn.softmax_quantizer TensorQuantizer(disabled) +model.language_model.layers.33.mlp.gate_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.33.mlp.gate_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.33.mlp.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.33.mlp.up_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.33.mlp.up_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.33.mlp.up_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.33.mlp.down_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.33.mlp.down_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.33.mlp.down_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.34.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.34.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.34.self_attn.q_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.34.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.34.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.34.self_attn.k_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.34.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.34.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.34.self_attn.v_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.34.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.34.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.34.self_attn.o_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.34.self_attn.q_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.34.self_attn.k_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.34.self_attn.v_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.34.self_attn.softmax_quantizer TensorQuantizer(disabled) +model.language_model.layers.34.mlp.gate_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.34.mlp.gate_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.34.mlp.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.34.mlp.up_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.34.mlp.up_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.34.mlp.up_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.34.mlp.down_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.34.mlp.down_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.34.mlp.down_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.35.self_attn.q_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.35.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.35.self_attn.q_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.35.self_attn.k_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.35.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.35.self_attn.k_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.35.self_attn.v_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.35.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.35.self_attn.v_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.35.self_attn.o_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.35.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.35.self_attn.o_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.35.self_attn.q_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.35.self_attn.k_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.35.self_attn.v_bmm_quantizer TensorQuantizer(disabled) +model.language_model.layers.35.self_attn.softmax_quantizer TensorQuantizer(disabled) +model.language_model.layers.35.mlp.gate_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.35.mlp.gate_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.35.mlp.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.35.mlp.up_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.35.mlp.up_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.35.mlp.up_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +model.language_model.layers.35.mlp.down_proj.input_quantizer TensorQuantizer(disabled) +model.language_model.layers.35.mlp.down_proj.output_quantizer TensorQuantizer(disabled) +model.language_model.layers.35.mlp.down_proj.weight_quantizer TensorQuantizer((2, 1) bit block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=255.0000 calibrator=MaxCalibrator quant) +lm_head.input_quantizer TensorQuantizer(disabled) +lm_head.output_quantizer TensorQuantizer(disabled) +lm_head.weight_quantizer TensorQuantizer(disabled) +1362 TensorQuantizers found in model