| { |
| "schema": "dynquant_checkpoint_v2", |
| "dynquant_core": "0.5.2", |
| "quantization_config": { |
| "quant_method": "dynquant", |
| "checkpoint_format": "dynquant-packed", |
| "schema_version": 1, |
| "version": "0.5.2", |
| "group_size": 128, |
| "symmetric": false, |
| "lm_head_quantized": true, |
| "modules_to_not_convert": [], |
| "modules": { |
| "lm_head": { |
| "bits": 8, |
| "out_features": 248320 |
| }, |
| "model.embed_tokens": { |
| "bits": 4, |
| "out_features": 248320 |
| }, |
| "model.layers.0.linear_attn.in_proj_a": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.0.linear_attn.in_proj_b": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.0.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "out_features": 10240 |
| }, |
| "model.layers.0.linear_attn.in_proj_z": { |
| "bits": 4, |
| "out_features": 6144 |
| }, |
| "model.layers.0.linear_attn.out_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.0.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.0.mlp.gate_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.0.mlp.up_proj": { |
| "bits": 2, |
| "out_features": 17408 |
| }, |
| "model.layers.1.linear_attn.in_proj_a": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.1.linear_attn.in_proj_b": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.1.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "out_features": 10240 |
| }, |
| "model.layers.1.linear_attn.in_proj_z": { |
| "bits": 4, |
| "out_features": 6144 |
| }, |
| "model.layers.1.linear_attn.out_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.1.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.1.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.1.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.10.linear_attn.in_proj_a": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.10.linear_attn.in_proj_b": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.10.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "out_features": 10240 |
| }, |
| "model.layers.10.linear_attn.in_proj_z": { |
| "bits": 4, |
| "out_features": 6144 |
| }, |
| "model.layers.10.linear_attn.out_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.10.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.10.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.10.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.11.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.11.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.11.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.11.self_attn.k_proj": { |
| "bits": 4, |
| "out_features": 1024 |
| }, |
| "model.layers.11.self_attn.o_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.11.self_attn.q_proj": { |
| "bits": 4, |
| "out_features": 12288 |
| }, |
| "model.layers.11.self_attn.v_proj": { |
| "bits": 4, |
| "out_features": 1024 |
| }, |
| "model.layers.12.linear_attn.in_proj_a": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.12.linear_attn.in_proj_b": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.12.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "out_features": 10240 |
| }, |
| "model.layers.12.linear_attn.in_proj_z": { |
| "bits": 4, |
| "out_features": 6144 |
| }, |
| "model.layers.12.linear_attn.out_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.12.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.12.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.12.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.13.linear_attn.in_proj_a": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.13.linear_attn.in_proj_b": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.13.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "out_features": 10240 |
| }, |
| "model.layers.13.linear_attn.in_proj_z": { |
| "bits": 4, |
| "out_features": 6144 |
| }, |
| "model.layers.13.linear_attn.out_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.13.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.13.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.13.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.14.linear_attn.in_proj_a": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.14.linear_attn.in_proj_b": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.14.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "out_features": 10240 |
| }, |
| "model.layers.14.linear_attn.in_proj_z": { |
| "bits": 4, |
| "out_features": 6144 |
| }, |
| "model.layers.14.linear_attn.out_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.14.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.14.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.14.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.15.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.15.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.15.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.15.self_attn.k_proj": { |
| "bits": 4, |
| "out_features": 1024 |
| }, |
| "model.layers.15.self_attn.o_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.15.self_attn.q_proj": { |
| "bits": 4, |
| "out_features": 12288 |
| }, |
| "model.layers.15.self_attn.v_proj": { |
| "bits": 4, |
| "out_features": 1024 |
| }, |
| "model.layers.16.linear_attn.in_proj_a": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.16.linear_attn.in_proj_b": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.16.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "out_features": 10240 |
| }, |
| "model.layers.16.linear_attn.in_proj_z": { |
| "bits": 4, |
| "out_features": 6144 |
| }, |
| "model.layers.16.linear_attn.out_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.16.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.16.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.16.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.17.linear_attn.in_proj_a": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.17.linear_attn.in_proj_b": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.17.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "out_features": 10240 |
| }, |
| "model.layers.17.linear_attn.in_proj_z": { |
| "bits": 4, |
| "out_features": 6144 |
| }, |
| "model.layers.17.linear_attn.out_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.17.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.17.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.17.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.18.linear_attn.in_proj_a": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.18.linear_attn.in_proj_b": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.18.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "out_features": 10240 |
| }, |
| "model.layers.18.linear_attn.in_proj_z": { |
| "bits": 4, |
| "out_features": 6144 |
| }, |
| "model.layers.18.linear_attn.out_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.18.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.18.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.18.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.19.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.19.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.19.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.19.self_attn.k_proj": { |
| "bits": 4, |
| "out_features": 1024 |
| }, |
| "model.layers.19.self_attn.o_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.19.self_attn.q_proj": { |
| "bits": 4, |
| "out_features": 12288 |
| }, |
| "model.layers.19.self_attn.v_proj": { |
| "bits": 4, |
| "out_features": 1024 |
| }, |
| "model.layers.2.linear_attn.in_proj_a": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.2.linear_attn.in_proj_b": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.2.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "out_features": 10240 |
| }, |
| "model.layers.2.linear_attn.in_proj_z": { |
| "bits": 4, |
| "out_features": 6144 |
| }, |
| "model.layers.2.linear_attn.out_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.2.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.2.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.2.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.20.linear_attn.in_proj_a": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.20.linear_attn.in_proj_b": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.20.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "out_features": 10240 |
| }, |
| "model.layers.20.linear_attn.in_proj_z": { |
| "bits": 4, |
| "out_features": 6144 |
| }, |
| "model.layers.20.linear_attn.out_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.20.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.20.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.20.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.21.linear_attn.in_proj_a": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.21.linear_attn.in_proj_b": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.21.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "out_features": 10240 |
| }, |
| "model.layers.21.linear_attn.in_proj_z": { |
| "bits": 4, |
| "out_features": 6144 |
| }, |
| "model.layers.21.linear_attn.out_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.21.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.21.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.21.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.22.linear_attn.in_proj_a": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.22.linear_attn.in_proj_b": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.22.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "out_features": 10240 |
| }, |
| "model.layers.22.linear_attn.in_proj_z": { |
| "bits": 4, |
| "out_features": 6144 |
| }, |
| "model.layers.22.linear_attn.out_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.22.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.22.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.22.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.23.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.23.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.23.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.23.self_attn.k_proj": { |
| "bits": 4, |
| "out_features": 1024 |
| }, |
| "model.layers.23.self_attn.o_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.23.self_attn.q_proj": { |
| "bits": 4, |
| "out_features": 12288 |
| }, |
| "model.layers.23.self_attn.v_proj": { |
| "bits": 4, |
| "out_features": 1024 |
| }, |
| "model.layers.24.linear_attn.in_proj_a": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.24.linear_attn.in_proj_b": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.24.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "out_features": 10240 |
| }, |
| "model.layers.24.linear_attn.in_proj_z": { |
| "bits": 4, |
| "out_features": 6144 |
| }, |
| "model.layers.24.linear_attn.out_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.24.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.24.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.24.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.25.linear_attn.in_proj_a": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.25.linear_attn.in_proj_b": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.25.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "out_features": 10240 |
| }, |
| "model.layers.25.linear_attn.in_proj_z": { |
| "bits": 4, |
| "out_features": 6144 |
| }, |
| "model.layers.25.linear_attn.out_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.25.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.25.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.25.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.26.linear_attn.in_proj_a": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.26.linear_attn.in_proj_b": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.26.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "out_features": 10240 |
| }, |
| "model.layers.26.linear_attn.in_proj_z": { |
| "bits": 4, |
| "out_features": 6144 |
| }, |
| "model.layers.26.linear_attn.out_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.26.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.26.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.26.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.27.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.27.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.27.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.27.self_attn.k_proj": { |
| "bits": 4, |
| "out_features": 1024 |
| }, |
| "model.layers.27.self_attn.o_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.27.self_attn.q_proj": { |
| "bits": 4, |
| "out_features": 12288 |
| }, |
| "model.layers.27.self_attn.v_proj": { |
| "bits": 4, |
| "out_features": 1024 |
| }, |
| "model.layers.28.linear_attn.in_proj_a": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.28.linear_attn.in_proj_b": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.28.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "out_features": 10240 |
| }, |
| "model.layers.28.linear_attn.in_proj_z": { |
| "bits": 4, |
| "out_features": 6144 |
| }, |
| "model.layers.28.linear_attn.out_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.28.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.28.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.28.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.29.linear_attn.in_proj_a": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.29.linear_attn.in_proj_b": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.29.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "out_features": 10240 |
| }, |
| "model.layers.29.linear_attn.in_proj_z": { |
| "bits": 4, |
| "out_features": 6144 |
| }, |
| "model.layers.29.linear_attn.out_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.29.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.29.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.29.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.3.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.3.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.3.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.3.self_attn.k_proj": { |
| "bits": 4, |
| "out_features": 1024 |
| }, |
| "model.layers.3.self_attn.o_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.3.self_attn.q_proj": { |
| "bits": 4, |
| "out_features": 12288 |
| }, |
| "model.layers.3.self_attn.v_proj": { |
| "bits": 4, |
| "out_features": 1024 |
| }, |
| "model.layers.30.linear_attn.in_proj_a": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.30.linear_attn.in_proj_b": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.30.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "out_features": 10240 |
| }, |
| "model.layers.30.linear_attn.in_proj_z": { |
| "bits": 4, |
| "out_features": 6144 |
| }, |
| "model.layers.30.linear_attn.out_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.30.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.30.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.30.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.31.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.31.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.31.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.31.self_attn.k_proj": { |
| "bits": 4, |
| "out_features": 1024 |
| }, |
| "model.layers.31.self_attn.o_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.31.self_attn.q_proj": { |
| "bits": 4, |
| "out_features": 12288 |
| }, |
| "model.layers.31.self_attn.v_proj": { |
| "bits": 4, |
| "out_features": 1024 |
| }, |
| "model.layers.32.linear_attn.in_proj_a": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.32.linear_attn.in_proj_b": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.32.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "out_features": 10240 |
| }, |
| "model.layers.32.linear_attn.in_proj_z": { |
| "bits": 4, |
| "out_features": 6144 |
| }, |
| "model.layers.32.linear_attn.out_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.32.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.32.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.32.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.33.linear_attn.in_proj_a": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.33.linear_attn.in_proj_b": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.33.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "out_features": 10240 |
| }, |
| "model.layers.33.linear_attn.in_proj_z": { |
| "bits": 4, |
| "out_features": 6144 |
| }, |
| "model.layers.33.linear_attn.out_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.33.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.33.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.33.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.34.linear_attn.in_proj_a": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.34.linear_attn.in_proj_b": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.34.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "out_features": 10240 |
| }, |
| "model.layers.34.linear_attn.in_proj_z": { |
| "bits": 4, |
| "out_features": 6144 |
| }, |
| "model.layers.34.linear_attn.out_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.34.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.34.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.34.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.35.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.35.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.35.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.35.self_attn.k_proj": { |
| "bits": 4, |
| "out_features": 1024 |
| }, |
| "model.layers.35.self_attn.o_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.35.self_attn.q_proj": { |
| "bits": 4, |
| "out_features": 12288 |
| }, |
| "model.layers.35.self_attn.v_proj": { |
| "bits": 4, |
| "out_features": 1024 |
| }, |
| "model.layers.36.linear_attn.in_proj_a": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.36.linear_attn.in_proj_b": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.36.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "out_features": 10240 |
| }, |
| "model.layers.36.linear_attn.in_proj_z": { |
| "bits": 4, |
| "out_features": 6144 |
| }, |
| "model.layers.36.linear_attn.out_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.36.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.36.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.36.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.37.linear_attn.in_proj_a": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.37.linear_attn.in_proj_b": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.37.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "out_features": 10240 |
| }, |
| "model.layers.37.linear_attn.in_proj_z": { |
| "bits": 4, |
| "out_features": 6144 |
| }, |
| "model.layers.37.linear_attn.out_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.37.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.37.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.37.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.38.linear_attn.in_proj_a": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.38.linear_attn.in_proj_b": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.38.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "out_features": 10240 |
| }, |
| "model.layers.38.linear_attn.in_proj_z": { |
| "bits": 4, |
| "out_features": 6144 |
| }, |
| "model.layers.38.linear_attn.out_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.38.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.38.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.38.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.39.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.39.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.39.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.39.self_attn.k_proj": { |
| "bits": 4, |
| "out_features": 1024 |
| }, |
| "model.layers.39.self_attn.o_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.39.self_attn.q_proj": { |
| "bits": 4, |
| "out_features": 12288 |
| }, |
| "model.layers.39.self_attn.v_proj": { |
| "bits": 4, |
| "out_features": 1024 |
| }, |
| "model.layers.4.linear_attn.in_proj_a": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.4.linear_attn.in_proj_b": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.4.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "out_features": 10240 |
| }, |
| "model.layers.4.linear_attn.in_proj_z": { |
| "bits": 4, |
| "out_features": 6144 |
| }, |
| "model.layers.4.linear_attn.out_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.4.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.4.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.4.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.40.linear_attn.in_proj_a": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.40.linear_attn.in_proj_b": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.40.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "out_features": 10240 |
| }, |
| "model.layers.40.linear_attn.in_proj_z": { |
| "bits": 4, |
| "out_features": 6144 |
| }, |
| "model.layers.40.linear_attn.out_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.40.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.40.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.40.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.41.linear_attn.in_proj_a": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.41.linear_attn.in_proj_b": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.41.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "out_features": 10240 |
| }, |
| "model.layers.41.linear_attn.in_proj_z": { |
| "bits": 4, |
| "out_features": 6144 |
| }, |
| "model.layers.41.linear_attn.out_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.41.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.41.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.41.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.42.linear_attn.in_proj_a": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.42.linear_attn.in_proj_b": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.42.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "out_features": 10240 |
| }, |
| "model.layers.42.linear_attn.in_proj_z": { |
| "bits": 4, |
| "out_features": 6144 |
| }, |
| "model.layers.42.linear_attn.out_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.42.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.42.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.42.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.43.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.43.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.43.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.43.self_attn.k_proj": { |
| "bits": 4, |
| "out_features": 1024 |
| }, |
| "model.layers.43.self_attn.o_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.43.self_attn.q_proj": { |
| "bits": 4, |
| "out_features": 12288 |
| }, |
| "model.layers.43.self_attn.v_proj": { |
| "bits": 4, |
| "out_features": 1024 |
| }, |
| "model.layers.44.linear_attn.in_proj_a": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.44.linear_attn.in_proj_b": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.44.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "out_features": 10240 |
| }, |
| "model.layers.44.linear_attn.in_proj_z": { |
| "bits": 4, |
| "out_features": 6144 |
| }, |
| "model.layers.44.linear_attn.out_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.44.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.44.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.44.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.45.linear_attn.in_proj_a": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.45.linear_attn.in_proj_b": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.45.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "out_features": 10240 |
| }, |
| "model.layers.45.linear_attn.in_proj_z": { |
| "bits": 4, |
| "out_features": 6144 |
| }, |
| "model.layers.45.linear_attn.out_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.45.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.45.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.45.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.46.linear_attn.in_proj_a": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.46.linear_attn.in_proj_b": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.46.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "out_features": 10240 |
| }, |
| "model.layers.46.linear_attn.in_proj_z": { |
| "bits": 4, |
| "out_features": 6144 |
| }, |
| "model.layers.46.linear_attn.out_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.46.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.46.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.46.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.47.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.47.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.47.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.47.self_attn.k_proj": { |
| "bits": 4, |
| "out_features": 1024 |
| }, |
| "model.layers.47.self_attn.o_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.47.self_attn.q_proj": { |
| "bits": 4, |
| "out_features": 12288 |
| }, |
| "model.layers.47.self_attn.v_proj": { |
| "bits": 4, |
| "out_features": 1024 |
| }, |
| "model.layers.48.linear_attn.in_proj_a": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.48.linear_attn.in_proj_b": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.48.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "out_features": 10240 |
| }, |
| "model.layers.48.linear_attn.in_proj_z": { |
| "bits": 4, |
| "out_features": 6144 |
| }, |
| "model.layers.48.linear_attn.out_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.48.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.48.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.48.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.49.linear_attn.in_proj_a": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.49.linear_attn.in_proj_b": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.49.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "out_features": 10240 |
| }, |
| "model.layers.49.linear_attn.in_proj_z": { |
| "bits": 4, |
| "out_features": 6144 |
| }, |
| "model.layers.49.linear_attn.out_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.49.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.49.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.49.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.5.linear_attn.in_proj_a": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.5.linear_attn.in_proj_b": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.5.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "out_features": 10240 |
| }, |
| "model.layers.5.linear_attn.in_proj_z": { |
| "bits": 4, |
| "out_features": 6144 |
| }, |
| "model.layers.5.linear_attn.out_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.5.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.5.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.5.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.50.linear_attn.in_proj_a": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.50.linear_attn.in_proj_b": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.50.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "out_features": 10240 |
| }, |
| "model.layers.50.linear_attn.in_proj_z": { |
| "bits": 4, |
| "out_features": 6144 |
| }, |
| "model.layers.50.linear_attn.out_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.50.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.50.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.50.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.51.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.51.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.51.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.51.self_attn.k_proj": { |
| "bits": 4, |
| "out_features": 1024 |
| }, |
| "model.layers.51.self_attn.o_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.51.self_attn.q_proj": { |
| "bits": 4, |
| "out_features": 12288 |
| }, |
| "model.layers.51.self_attn.v_proj": { |
| "bits": 4, |
| "out_features": 1024 |
| }, |
| "model.layers.52.linear_attn.in_proj_a": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.52.linear_attn.in_proj_b": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.52.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "out_features": 10240 |
| }, |
| "model.layers.52.linear_attn.in_proj_z": { |
| "bits": 4, |
| "out_features": 6144 |
| }, |
| "model.layers.52.linear_attn.out_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.52.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.52.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.52.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.53.linear_attn.in_proj_a": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.53.linear_attn.in_proj_b": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.53.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "out_features": 10240 |
| }, |
| "model.layers.53.linear_attn.in_proj_z": { |
| "bits": 4, |
| "out_features": 6144 |
| }, |
| "model.layers.53.linear_attn.out_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.53.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.53.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.53.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.54.linear_attn.in_proj_a": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.54.linear_attn.in_proj_b": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.54.linear_attn.in_proj_qkv": { |
| "bits": 3, |
| "out_features": 10240 |
| }, |
| "model.layers.54.linear_attn.in_proj_z": { |
| "bits": 4, |
| "out_features": 6144 |
| }, |
| "model.layers.54.linear_attn.out_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.54.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.54.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.54.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.55.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.55.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.55.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.55.self_attn.k_proj": { |
| "bits": 4, |
| "out_features": 1024 |
| }, |
| "model.layers.55.self_attn.o_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.55.self_attn.q_proj": { |
| "bits": 3, |
| "out_features": 12288 |
| }, |
| "model.layers.55.self_attn.v_proj": { |
| "bits": 4, |
| "out_features": 1024 |
| }, |
| "model.layers.56.linear_attn.in_proj_a": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.56.linear_attn.in_proj_b": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.56.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "out_features": 10240 |
| }, |
| "model.layers.56.linear_attn.in_proj_z": { |
| "bits": 4, |
| "out_features": 6144 |
| }, |
| "model.layers.56.linear_attn.out_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.56.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.56.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.56.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.57.linear_attn.in_proj_a": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.57.linear_attn.in_proj_b": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.57.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "out_features": 10240 |
| }, |
| "model.layers.57.linear_attn.in_proj_z": { |
| "bits": 4, |
| "out_features": 6144 |
| }, |
| "model.layers.57.linear_attn.out_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.57.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.57.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.57.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.58.linear_attn.in_proj_a": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.58.linear_attn.in_proj_b": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.58.linear_attn.in_proj_qkv": { |
| "bits": 3, |
| "out_features": 10240 |
| }, |
| "model.layers.58.linear_attn.in_proj_z": { |
| "bits": 3, |
| "out_features": 6144 |
| }, |
| "model.layers.58.linear_attn.out_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.58.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.58.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.58.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.59.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.59.mlp.gate_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.59.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.59.self_attn.k_proj": { |
| "bits": 4, |
| "out_features": 1024 |
| }, |
| "model.layers.59.self_attn.o_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.59.self_attn.q_proj": { |
| "bits": 4, |
| "out_features": 12288 |
| }, |
| "model.layers.59.self_attn.v_proj": { |
| "bits": 4, |
| "out_features": 1024 |
| }, |
| "model.layers.6.linear_attn.in_proj_a": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.6.linear_attn.in_proj_b": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.6.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "out_features": 10240 |
| }, |
| "model.layers.6.linear_attn.in_proj_z": { |
| "bits": 4, |
| "out_features": 6144 |
| }, |
| "model.layers.6.linear_attn.out_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.6.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.6.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.6.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.60.linear_attn.in_proj_a": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.60.linear_attn.in_proj_b": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.60.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "out_features": 10240 |
| }, |
| "model.layers.60.linear_attn.in_proj_z": { |
| "bits": 4, |
| "out_features": 6144 |
| }, |
| "model.layers.60.linear_attn.out_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.60.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.60.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.60.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.61.linear_attn.in_proj_a": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.61.linear_attn.in_proj_b": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.61.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "out_features": 10240 |
| }, |
| "model.layers.61.linear_attn.in_proj_z": { |
| "bits": 4, |
| "out_features": 6144 |
| }, |
| "model.layers.61.linear_attn.out_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.61.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.61.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.61.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.62.linear_attn.in_proj_a": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.62.linear_attn.in_proj_b": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.62.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "out_features": 10240 |
| }, |
| "model.layers.62.linear_attn.in_proj_z": { |
| "bits": 4, |
| "out_features": 6144 |
| }, |
| "model.layers.62.linear_attn.out_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.62.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.62.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.62.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.63.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.63.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.63.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.63.self_attn.k_proj": { |
| "bits": 4, |
| "out_features": 1024 |
| }, |
| "model.layers.63.self_attn.o_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.63.self_attn.q_proj": { |
| "bits": 4, |
| "out_features": 12288 |
| }, |
| "model.layers.63.self_attn.v_proj": { |
| "bits": 4, |
| "out_features": 1024 |
| }, |
| "model.layers.7.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.7.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.7.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.7.self_attn.k_proj": { |
| "bits": 4, |
| "out_features": 1024 |
| }, |
| "model.layers.7.self_attn.o_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.7.self_attn.q_proj": { |
| "bits": 4, |
| "out_features": 12288 |
| }, |
| "model.layers.7.self_attn.v_proj": { |
| "bits": 4, |
| "out_features": 1024 |
| }, |
| "model.layers.8.linear_attn.in_proj_a": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.8.linear_attn.in_proj_b": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.8.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "out_features": 10240 |
| }, |
| "model.layers.8.linear_attn.in_proj_z": { |
| "bits": 4, |
| "out_features": 6144 |
| }, |
| "model.layers.8.linear_attn.out_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.8.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.8.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.8.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.9.linear_attn.in_proj_a": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.9.linear_attn.in_proj_b": { |
| "bits": 8, |
| "out_features": 48 |
| }, |
| "model.layers.9.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "out_features": 10240 |
| }, |
| "model.layers.9.linear_attn.in_proj_z": { |
| "bits": 4, |
| "out_features": 6144 |
| }, |
| "model.layers.9.linear_attn.out_proj": { |
| "bits": 4, |
| "out_features": 5120 |
| }, |
| "model.layers.9.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.9.mlp.gate_proj": { |
| "bits": 4, |
| "out_features": 17408 |
| }, |
| "model.layers.9.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| } |
| } |
| }, |
| "files": [ |
| "model-00001-of-00004.safetensors", |
| "model-00002-of-00004.safetensors", |
| "model-00003-of-00004.safetensors", |
| "model-00004-of-00004.safetensors", |
| "model.safetensors.index.json" |
| ], |
| "accounting": { |
| "packed_nbytes": 13442334720, |
| "dense_nbytes": 5291008, |
| "total_nbytes": 13447625728, |
| "quantized_elements": 26893352960, |
| "average_bits": 3.998708 |
| }, |
| "tied": {}, |
| "expert_banks": [], |
| "provenance": { |
| "dynquant_core": "0.5.2", |
| "model": "/workspace/runs-s2/qwen38-27b.text2sql/merged", |
| "stats": "/workspace/runs-s2/qwen38-27b.text2sql/stats/dynquant_stats.json", |
| "allocator": "rank_product", |
| "widths_from": "4.00 avg bits", |
| "predicted_nbytes": 13447625728 |
| }, |
| "layers": { |
| "lm_head": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 248320, |
| 5120 |
| ], |
| "num_rows": 248320, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 1311129600, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.embed_tokens": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 248320, |
| 5120 |
| ], |
| "num_rows": 248320, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 675430400, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.953984, |
| "clip_improvement": 0.112084 |
| }, |
| "model.layers.0.linear_attn.in_proj_a": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.0.linear_attn.in_proj_b": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.0.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 27852800, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.95415, |
| "clip_improvement": 0.111904 |
| }, |
| "model.layers.0.linear_attn.in_proj_z": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954659, |
| "clip_improvement": 0.111781 |
| }, |
| "model.layers.0.linear_attn.out_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.956059, |
| "clip_improvement": 0.110666 |
| }, |
| "model.layers.0.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999849, |
| "clip_improvement": 0.250686 |
| }, |
| "model.layers.0.mlp.gate_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999854, |
| "clip_improvement": 0.250003 |
| }, |
| "model.layers.0.mlp.up_proj": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 25067520, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.999978, |
| "clip_improvement": 0.353266 |
| }, |
| "model.layers.1.linear_attn.in_proj_a": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.1.linear_attn.in_proj_b": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.1.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 27852800, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954595, |
| "clip_improvement": 0.112052 |
| }, |
| "model.layers.1.linear_attn.in_proj_z": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954106, |
| "clip_improvement": 0.111822 |
| }, |
| "model.layers.1.linear_attn.out_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954895, |
| "clip_improvement": 0.111771 |
| }, |
| "model.layers.1.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999822, |
| "clip_improvement": 0.251641 |
| }, |
| "model.layers.1.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.953819, |
| "clip_improvement": 0.111691 |
| }, |
| "model.layers.1.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999848, |
| "clip_improvement": 0.249995 |
| }, |
| "model.layers.10.linear_attn.in_proj_a": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.10.linear_attn.in_proj_b": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.10.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 27852800, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955293, |
| "clip_improvement": 0.111847 |
| }, |
| "model.layers.10.linear_attn.in_proj_z": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.95544, |
| "clip_improvement": 0.111705 |
| }, |
| "model.layers.10.linear_attn.out_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955514, |
| "clip_improvement": 0.111696 |
| }, |
| "model.layers.10.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999858, |
| "clip_improvement": 0.250209 |
| }, |
| "model.layers.10.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.95357, |
| "clip_improvement": 0.111593 |
| }, |
| "model.layers.10.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999848, |
| "clip_improvement": 0.250044 |
| }, |
| "model.layers.11.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999842, |
| "clip_improvement": 0.250375 |
| }, |
| "model.layers.11.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954025, |
| "clip_improvement": 0.111565 |
| }, |
| "model.layers.11.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999866, |
| "clip_improvement": 0.24994 |
| }, |
| "model.layers.11.self_attn.k_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 2785280, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.956763, |
| "clip_improvement": 0.111992 |
| }, |
| "model.layers.11.self_attn.o_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954488, |
| "clip_improvement": 0.111156 |
| }, |
| "model.layers.11.self_attn.q_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 12288, |
| 5120 |
| ], |
| "num_rows": 12288, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 33423360, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955436, |
| "clip_improvement": 0.111732 |
| }, |
| "model.layers.11.self_attn.v_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 2785280, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.95564, |
| "clip_improvement": 0.111539 |
| }, |
| "model.layers.12.linear_attn.in_proj_a": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.12.linear_attn.in_proj_b": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.12.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 27852800, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954878, |
| "clip_improvement": 0.111612 |
| }, |
| "model.layers.12.linear_attn.in_proj_z": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955335, |
| "clip_improvement": 0.111938 |
| }, |
| "model.layers.12.linear_attn.out_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954671, |
| "clip_improvement": 0.111786 |
| }, |
| "model.layers.12.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999842, |
| "clip_improvement": 0.250405 |
| }, |
| "model.layers.12.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.953875, |
| "clip_improvement": 0.111618 |
| }, |
| "model.layers.12.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999832, |
| "clip_improvement": 0.249959 |
| }, |
| "model.layers.13.linear_attn.in_proj_a": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.13.linear_attn.in_proj_b": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.13.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 27852800, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955083, |
| "clip_improvement": 0.111819 |
| }, |
| "model.layers.13.linear_attn.in_proj_z": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954443, |
| "clip_improvement": 0.111811 |
| }, |
| "model.layers.13.linear_attn.out_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955811, |
| "clip_improvement": 0.111977 |
| }, |
| "model.layers.13.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999806, |
| "clip_improvement": 0.249997 |
| }, |
| "model.layers.13.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954267, |
| "clip_improvement": 0.111558 |
| }, |
| "model.layers.13.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999848, |
| "clip_improvement": 0.25019 |
| }, |
| "model.layers.14.linear_attn.in_proj_a": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.14.linear_attn.in_proj_b": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.14.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 27852800, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955312, |
| "clip_improvement": 0.111958 |
| }, |
| "model.layers.14.linear_attn.in_proj_z": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954594, |
| "clip_improvement": 0.112234 |
| }, |
| "model.layers.14.linear_attn.out_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955127, |
| "clip_improvement": 0.111368 |
| }, |
| "model.layers.14.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999855, |
| "clip_improvement": 0.250201 |
| }, |
| "model.layers.14.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.95388, |
| "clip_improvement": 0.111457 |
| }, |
| "model.layers.14.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999858, |
| "clip_improvement": 0.250083 |
| }, |
| "model.layers.15.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999876, |
| "clip_improvement": 0.250315 |
| }, |
| "model.layers.15.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954024, |
| "clip_improvement": 0.111581 |
| }, |
| "model.layers.15.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999876, |
| "clip_improvement": 0.25008 |
| }, |
| "model.layers.15.self_attn.k_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 2785280, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955322, |
| "clip_improvement": 0.111896 |
| }, |
| "model.layers.15.self_attn.o_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954944, |
| "clip_improvement": 0.111918 |
| }, |
| "model.layers.15.self_attn.q_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 12288, |
| 5120 |
| ], |
| "num_rows": 12288, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 33423360, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954852, |
| "clip_improvement": 0.112017 |
| }, |
| "model.layers.15.self_attn.v_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 2785280, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954248, |
| "clip_improvement": 0.111325 |
| }, |
| "model.layers.16.linear_attn.in_proj_a": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.16.linear_attn.in_proj_b": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.16.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 27852800, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954963, |
| "clip_improvement": 0.111801 |
| }, |
| "model.layers.16.linear_attn.in_proj_z": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954997, |
| "clip_improvement": 0.111832 |
| }, |
| "model.layers.16.linear_attn.out_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.95426, |
| "clip_improvement": 0.111103 |
| }, |
| "model.layers.16.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999854, |
| "clip_improvement": 0.250263 |
| }, |
| "model.layers.16.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.953856, |
| "clip_improvement": 0.111647 |
| }, |
| "model.layers.16.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999862, |
| "clip_improvement": 0.250095 |
| }, |
| "model.layers.17.linear_attn.in_proj_a": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.17.linear_attn.in_proj_b": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.17.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 27852800, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955305, |
| "clip_improvement": 0.111751 |
| }, |
| "model.layers.17.linear_attn.in_proj_z": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954956, |
| "clip_improvement": 0.112033 |
| }, |
| "model.layers.17.linear_attn.out_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955562, |
| "clip_improvement": 0.111654 |
| }, |
| "model.layers.17.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999862, |
| "clip_improvement": 0.250546 |
| }, |
| "model.layers.17.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.953981, |
| "clip_improvement": 0.111532 |
| }, |
| "model.layers.17.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999826, |
| "clip_improvement": 0.250273 |
| }, |
| "model.layers.18.linear_attn.in_proj_a": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.18.linear_attn.in_proj_b": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.18.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 27852800, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955105, |
| "clip_improvement": 0.111719 |
| }, |
| "model.layers.18.linear_attn.in_proj_z": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955725, |
| "clip_improvement": 0.111969 |
| }, |
| "model.layers.18.linear_attn.out_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955257, |
| "clip_improvement": 0.112642 |
| }, |
| "model.layers.18.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999642, |
| "clip_improvement": 0.254122 |
| }, |
| "model.layers.18.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954015, |
| "clip_improvement": 0.111616 |
| }, |
| "model.layers.18.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999813, |
| "clip_improvement": 0.250417 |
| }, |
| "model.layers.19.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999861, |
| "clip_improvement": 0.250504 |
| }, |
| "model.layers.19.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.95399, |
| "clip_improvement": 0.111614 |
| }, |
| "model.layers.19.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999841, |
| "clip_improvement": 0.250121 |
| }, |
| "model.layers.19.self_attn.k_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 2785280, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.956494, |
| "clip_improvement": 0.111651 |
| }, |
| "model.layers.19.self_attn.o_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.95542, |
| "clip_improvement": 0.111117 |
| }, |
| "model.layers.19.self_attn.q_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 12288, |
| 5120 |
| ], |
| "num_rows": 12288, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 33423360, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954928, |
| "clip_improvement": 0.111848 |
| }, |
| "model.layers.19.self_attn.v_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 2785280, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.958228, |
| "clip_improvement": 0.112204 |
| }, |
| "model.layers.2.linear_attn.in_proj_a": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.2.linear_attn.in_proj_b": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.2.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 27852800, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.95467, |
| "clip_improvement": 0.111766 |
| }, |
| "model.layers.2.linear_attn.in_proj_z": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.95437, |
| "clip_improvement": 0.11178 |
| }, |
| "model.layers.2.linear_attn.out_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955103, |
| "clip_improvement": 0.111672 |
| }, |
| "model.layers.2.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999854, |
| "clip_improvement": 0.250175 |
| }, |
| "model.layers.2.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.953527, |
| "clip_improvement": 0.111621 |
| }, |
| "model.layers.2.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999851, |
| "clip_improvement": 0.250174 |
| }, |
| "model.layers.20.linear_attn.in_proj_a": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.20.linear_attn.in_proj_b": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.20.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 27852800, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.95561, |
| "clip_improvement": 0.111642 |
| }, |
| "model.layers.20.linear_attn.in_proj_z": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954899, |
| "clip_improvement": 0.111846 |
| }, |
| "model.layers.20.linear_attn.out_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.956291, |
| "clip_improvement": 0.111761 |
| }, |
| "model.layers.20.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999859, |
| "clip_improvement": 0.250298 |
| }, |
| "model.layers.20.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954209, |
| "clip_improvement": 0.111617 |
| }, |
| "model.layers.20.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999849, |
| "clip_improvement": 0.250163 |
| }, |
| "model.layers.21.linear_attn.in_proj_a": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.21.linear_attn.in_proj_b": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.21.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 27852800, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.9551, |
| "clip_improvement": 0.11174 |
| }, |
| "model.layers.21.linear_attn.in_proj_z": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954683, |
| "clip_improvement": 0.111484 |
| }, |
| "model.layers.21.linear_attn.out_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955351, |
| "clip_improvement": 0.111443 |
| }, |
| "model.layers.21.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999832, |
| "clip_improvement": 0.25046 |
| }, |
| "model.layers.21.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.95384, |
| "clip_improvement": 0.111601 |
| }, |
| "model.layers.21.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999833, |
| "clip_improvement": 0.250088 |
| }, |
| "model.layers.22.linear_attn.in_proj_a": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.22.linear_attn.in_proj_b": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.22.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 27852800, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954575, |
| "clip_improvement": 0.111636 |
| }, |
| "model.layers.22.linear_attn.in_proj_z": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.95437, |
| "clip_improvement": 0.111671 |
| }, |
| "model.layers.22.linear_attn.out_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955615, |
| "clip_improvement": 0.11214 |
| }, |
| "model.layers.22.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999658, |
| "clip_improvement": 0.254971 |
| }, |
| "model.layers.22.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954316, |
| "clip_improvement": 0.111766 |
| }, |
| "model.layers.22.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999843, |
| "clip_improvement": 0.250311 |
| }, |
| "model.layers.23.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999876, |
| "clip_improvement": 0.250967 |
| }, |
| "model.layers.23.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954176, |
| "clip_improvement": 0.11172 |
| }, |
| "model.layers.23.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999836, |
| "clip_improvement": 0.250167 |
| }, |
| "model.layers.23.self_attn.k_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 2785280, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.956104, |
| "clip_improvement": 0.111634 |
| }, |
| "model.layers.23.self_attn.o_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954574, |
| "clip_improvement": 0.111474 |
| }, |
| "model.layers.23.self_attn.q_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 12288, |
| 5120 |
| ], |
| "num_rows": 12288, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 33423360, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955151, |
| "clip_improvement": 0.11184 |
| }, |
| "model.layers.23.self_attn.v_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 2785280, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.956299, |
| "clip_improvement": 0.111033 |
| }, |
| "model.layers.24.linear_attn.in_proj_a": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.24.linear_attn.in_proj_b": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.24.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 27852800, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955276, |
| "clip_improvement": 0.111569 |
| }, |
| "model.layers.24.linear_attn.in_proj_z": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955094, |
| "clip_improvement": 0.11148 |
| }, |
| "model.layers.24.linear_attn.out_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954879, |
| "clip_improvement": 0.111599 |
| }, |
| "model.layers.24.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999838, |
| "clip_improvement": 0.250928 |
| }, |
| "model.layers.24.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954313, |
| "clip_improvement": 0.111565 |
| }, |
| "model.layers.24.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999833, |
| "clip_improvement": 0.250264 |
| }, |
| "model.layers.25.linear_attn.in_proj_a": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.25.linear_attn.in_proj_b": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.25.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 27852800, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.95554, |
| "clip_improvement": 0.111571 |
| }, |
| "model.layers.25.linear_attn.in_proj_z": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955497, |
| "clip_improvement": 0.111745 |
| }, |
| "model.layers.25.linear_attn.out_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954545, |
| "clip_improvement": 0.111737 |
| }, |
| "model.layers.25.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999856, |
| "clip_improvement": 0.250225 |
| }, |
| "model.layers.25.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954248, |
| "clip_improvement": 0.11155 |
| }, |
| "model.layers.25.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999835, |
| "clip_improvement": 0.250287 |
| }, |
| "model.layers.26.linear_attn.in_proj_a": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.26.linear_attn.in_proj_b": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.26.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 27852800, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955549, |
| "clip_improvement": 0.11143 |
| }, |
| "model.layers.26.linear_attn.in_proj_z": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954757, |
| "clip_improvement": 0.111722 |
| }, |
| "model.layers.26.linear_attn.out_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954948, |
| "clip_improvement": 0.111902 |
| }, |
| "model.layers.26.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999845, |
| "clip_improvement": 0.250635 |
| }, |
| "model.layers.26.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954525, |
| "clip_improvement": 0.111576 |
| }, |
| "model.layers.26.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999866, |
| "clip_improvement": 0.250615 |
| }, |
| "model.layers.27.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999855, |
| "clip_improvement": 0.250789 |
| }, |
| "model.layers.27.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954277, |
| "clip_improvement": 0.111582 |
| }, |
| "model.layers.27.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999861, |
| "clip_improvement": 0.250364 |
| }, |
| "model.layers.27.self_attn.k_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 2785280, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.958398, |
| "clip_improvement": 0.111747 |
| }, |
| "model.layers.27.self_attn.o_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955628, |
| "clip_improvement": 0.111578 |
| }, |
| "model.layers.27.self_attn.q_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 12288, |
| 5120 |
| ], |
| "num_rows": 12288, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 33423360, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954954, |
| "clip_improvement": 0.111449 |
| }, |
| "model.layers.27.self_attn.v_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 2785280, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.956763, |
| "clip_improvement": 0.110793 |
| }, |
| "model.layers.28.linear_attn.in_proj_a": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.28.linear_attn.in_proj_b": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.28.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 27852800, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.957024, |
| "clip_improvement": 0.111722 |
| }, |
| "model.layers.28.linear_attn.in_proj_z": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.95516, |
| "clip_improvement": 0.111571 |
| }, |
| "model.layers.28.linear_attn.out_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955054, |
| "clip_improvement": 0.111638 |
| }, |
| "model.layers.28.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999862, |
| "clip_improvement": 0.250542 |
| }, |
| "model.layers.28.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954699, |
| "clip_improvement": 0.111631 |
| }, |
| "model.layers.28.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999864, |
| "clip_improvement": 0.250489 |
| }, |
| "model.layers.29.linear_attn.in_proj_a": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.29.linear_attn.in_proj_b": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.29.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 27852800, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955435, |
| "clip_improvement": 0.111626 |
| }, |
| "model.layers.29.linear_attn.in_proj_z": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955644, |
| "clip_improvement": 0.111821 |
| }, |
| "model.layers.29.linear_attn.out_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955623, |
| "clip_improvement": 0.111556 |
| }, |
| "model.layers.29.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999856, |
| "clip_improvement": 0.250275 |
| }, |
| "model.layers.29.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954318, |
| "clip_improvement": 0.111557 |
| }, |
| "model.layers.29.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999838, |
| "clip_improvement": 0.250411 |
| }, |
| "model.layers.3.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999836, |
| "clip_improvement": 0.250541 |
| }, |
| "model.layers.3.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.953631, |
| "clip_improvement": 0.111623 |
| }, |
| "model.layers.3.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999838, |
| "clip_improvement": 0.249902 |
| }, |
| "model.layers.3.self_attn.k_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 2785280, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954395, |
| "clip_improvement": 0.111428 |
| }, |
| "model.layers.3.self_attn.o_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.956112, |
| "clip_improvement": 0.112121 |
| }, |
| "model.layers.3.self_attn.q_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 12288, |
| 5120 |
| ], |
| "num_rows": 12288, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 33423360, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954397, |
| "clip_improvement": 0.111928 |
| }, |
| "model.layers.3.self_attn.v_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 2785280, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954688, |
| "clip_improvement": 0.111859 |
| }, |
| "model.layers.30.linear_attn.in_proj_a": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.30.linear_attn.in_proj_b": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.30.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 27852800, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.956084, |
| "clip_improvement": 0.111422 |
| }, |
| "model.layers.30.linear_attn.in_proj_z": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955347, |
| "clip_improvement": 0.11186 |
| }, |
| "model.layers.30.linear_attn.out_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.95518, |
| "clip_improvement": 0.111644 |
| }, |
| "model.layers.30.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999835, |
| "clip_improvement": 0.250864 |
| }, |
| "model.layers.30.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954779, |
| "clip_improvement": 0.111442 |
| }, |
| "model.layers.30.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999848, |
| "clip_improvement": 0.25042 |
| }, |
| "model.layers.31.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999815, |
| "clip_improvement": 0.250486 |
| }, |
| "model.layers.31.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954723, |
| "clip_improvement": 0.111373 |
| }, |
| "model.layers.31.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999854, |
| "clip_improvement": 0.250653 |
| }, |
| "model.layers.31.self_attn.k_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 2785280, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.958081, |
| "clip_improvement": 0.111609 |
| }, |
| "model.layers.31.self_attn.o_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954903, |
| "clip_improvement": 0.111542 |
| }, |
| "model.layers.31.self_attn.q_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 12288, |
| 5120 |
| ], |
| "num_rows": 12288, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 33423360, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955337, |
| "clip_improvement": 0.111529 |
| }, |
| "model.layers.31.self_attn.v_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 2785280, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.958643, |
| "clip_improvement": 0.110719 |
| }, |
| "model.layers.32.linear_attn.in_proj_a": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.32.linear_attn.in_proj_b": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.32.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 27852800, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955952, |
| "clip_improvement": 0.111278 |
| }, |
| "model.layers.32.linear_attn.in_proj_z": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955221, |
| "clip_improvement": 0.111688 |
| }, |
| "model.layers.32.linear_attn.out_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954883, |
| "clip_improvement": 0.111769 |
| }, |
| "model.layers.32.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999836, |
| "clip_improvement": 0.250786 |
| }, |
| "model.layers.32.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954603, |
| "clip_improvement": 0.111334 |
| }, |
| "model.layers.32.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999849, |
| "clip_improvement": 0.250684 |
| }, |
| "model.layers.33.linear_attn.in_proj_a": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.33.linear_attn.in_proj_b": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.33.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 27852800, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.956208, |
| "clip_improvement": 0.11136 |
| }, |
| "model.layers.33.linear_attn.in_proj_z": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955294, |
| "clip_improvement": 0.112067 |
| }, |
| "model.layers.33.linear_attn.out_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955444, |
| "clip_improvement": 0.11159 |
| }, |
| "model.layers.33.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999859, |
| "clip_improvement": 0.250823 |
| }, |
| "model.layers.33.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954881, |
| "clip_improvement": 0.111475 |
| }, |
| "model.layers.33.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999835, |
| "clip_improvement": 0.250689 |
| }, |
| "model.layers.34.linear_attn.in_proj_a": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.34.linear_attn.in_proj_b": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.34.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 27852800, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955842, |
| "clip_improvement": 0.111378 |
| }, |
| "model.layers.34.linear_attn.in_proj_z": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954972, |
| "clip_improvement": 0.112012 |
| }, |
| "model.layers.34.linear_attn.out_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955933, |
| "clip_improvement": 0.111223 |
| }, |
| "model.layers.34.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999698, |
| "clip_improvement": 0.252964 |
| }, |
| "model.layers.34.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955424, |
| "clip_improvement": 0.111511 |
| }, |
| "model.layers.34.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999839, |
| "clip_improvement": 0.250805 |
| }, |
| "model.layers.35.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999854, |
| "clip_improvement": 0.250783 |
| }, |
| "model.layers.35.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954238, |
| "clip_improvement": 0.11157 |
| }, |
| "model.layers.35.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999845, |
| "clip_improvement": 0.250573 |
| }, |
| "model.layers.35.self_attn.k_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 2785280, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955542, |
| "clip_improvement": 0.111618 |
| }, |
| "model.layers.35.self_attn.o_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955241, |
| "clip_improvement": 0.111158 |
| }, |
| "model.layers.35.self_attn.q_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 12288, |
| 5120 |
| ], |
| "num_rows": 12288, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 33423360, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955658, |
| "clip_improvement": 0.111613 |
| }, |
| "model.layers.35.self_attn.v_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 2785280, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.957959, |
| "clip_improvement": 0.11173 |
| }, |
| "model.layers.36.linear_attn.in_proj_a": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.36.linear_attn.in_proj_b": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.36.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 27852800, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955959, |
| "clip_improvement": 0.111665 |
| }, |
| "model.layers.36.linear_attn.in_proj_z": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954822, |
| "clip_improvement": 0.111682 |
| }, |
| "model.layers.36.linear_attn.out_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.956401, |
| "clip_improvement": 0.111882 |
| }, |
| "model.layers.36.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999832, |
| "clip_improvement": 0.250335 |
| }, |
| "model.layers.36.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.953918, |
| "clip_improvement": 0.111382 |
| }, |
| "model.layers.36.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999875, |
| "clip_improvement": 0.250406 |
| }, |
| "model.layers.37.linear_attn.in_proj_a": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.37.linear_attn.in_proj_b": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.37.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 27852800, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955498, |
| "clip_improvement": 0.111732 |
| }, |
| "model.layers.37.linear_attn.in_proj_z": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955689, |
| "clip_improvement": 0.111617 |
| }, |
| "model.layers.37.linear_attn.out_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.956165, |
| "clip_improvement": 0.111772 |
| }, |
| "model.layers.37.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999848, |
| "clip_improvement": 0.250149 |
| }, |
| "model.layers.37.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954048, |
| "clip_improvement": 0.11159 |
| }, |
| "model.layers.37.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999835, |
| "clip_improvement": 0.25028 |
| }, |
| "model.layers.38.linear_attn.in_proj_a": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.38.linear_attn.in_proj_b": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.38.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 27852800, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.95522, |
| "clip_improvement": 0.111837 |
| }, |
| "model.layers.38.linear_attn.in_proj_z": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954639, |
| "clip_improvement": 0.111934 |
| }, |
| "model.layers.38.linear_attn.out_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955221, |
| "clip_improvement": 0.111493 |
| }, |
| "model.layers.38.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999841, |
| "clip_improvement": 0.250749 |
| }, |
| "model.layers.38.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954482, |
| "clip_improvement": 0.111663 |
| }, |
| "model.layers.38.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999856, |
| "clip_improvement": 0.250307 |
| }, |
| "model.layers.39.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999881, |
| "clip_improvement": 0.250513 |
| }, |
| "model.layers.39.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954291, |
| "clip_improvement": 0.11161 |
| }, |
| "model.layers.39.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999864, |
| "clip_improvement": 0.250214 |
| }, |
| "model.layers.39.self_attn.k_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 2785280, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955273, |
| "clip_improvement": 0.111974 |
| }, |
| "model.layers.39.self_attn.o_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954614, |
| "clip_improvement": 0.111142 |
| }, |
| "model.layers.39.self_attn.q_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 12288, |
| 5120 |
| ], |
| "num_rows": 12288, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 33423360, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954856, |
| "clip_improvement": 0.111827 |
| }, |
| "model.layers.39.self_attn.v_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 2785280, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.957812, |
| "clip_improvement": 0.111052 |
| }, |
| "model.layers.4.linear_attn.in_proj_a": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.4.linear_attn.in_proj_b": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.4.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 27852800, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954687, |
| "clip_improvement": 0.111505 |
| }, |
| "model.layers.4.linear_attn.in_proj_z": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955282, |
| "clip_improvement": 0.112024 |
| }, |
| "model.layers.4.linear_attn.out_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954976, |
| "clip_improvement": 0.110724 |
| }, |
| "model.layers.4.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999838, |
| "clip_improvement": 0.250218 |
| }, |
| "model.layers.4.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.953931, |
| "clip_improvement": 0.111602 |
| }, |
| "model.layers.4.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999865, |
| "clip_improvement": 0.25003 |
| }, |
| "model.layers.40.linear_attn.in_proj_a": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.40.linear_attn.in_proj_b": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.40.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 27852800, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955464, |
| "clip_improvement": 0.111769 |
| }, |
| "model.layers.40.linear_attn.in_proj_z": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954598, |
| "clip_improvement": 0.111617 |
| }, |
| "model.layers.40.linear_attn.out_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955359, |
| "clip_improvement": 0.111777 |
| }, |
| "model.layers.40.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999848, |
| "clip_improvement": 0.250317 |
| }, |
| "model.layers.40.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954341, |
| "clip_improvement": 0.111629 |
| }, |
| "model.layers.40.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999864, |
| "clip_improvement": 0.250118 |
| }, |
| "model.layers.41.linear_attn.in_proj_a": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.41.linear_attn.in_proj_b": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.41.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 27852800, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955395, |
| "clip_improvement": 0.111947 |
| }, |
| "model.layers.41.linear_attn.in_proj_z": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955819, |
| "clip_improvement": 0.111906 |
| }, |
| "model.layers.41.linear_attn.out_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955778, |
| "clip_improvement": 0.111758 |
| }, |
| "model.layers.41.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999866, |
| "clip_improvement": 0.250512 |
| }, |
| "model.layers.41.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954525, |
| "clip_improvement": 0.111461 |
| }, |
| "model.layers.41.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999875, |
| "clip_improvement": 0.250346 |
| }, |
| "model.layers.42.linear_attn.in_proj_a": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.42.linear_attn.in_proj_b": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.42.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 27852800, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.956357, |
| "clip_improvement": 0.111743 |
| }, |
| "model.layers.42.linear_attn.in_proj_z": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.95507, |
| "clip_improvement": 0.111862 |
| }, |
| "model.layers.42.linear_attn.out_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955485, |
| "clip_improvement": 0.111715 |
| }, |
| "model.layers.42.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999835, |
| "clip_improvement": 0.250647 |
| }, |
| "model.layers.42.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954532, |
| "clip_improvement": 0.111599 |
| }, |
| "model.layers.42.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999868, |
| "clip_improvement": 0.250664 |
| }, |
| "model.layers.43.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999843, |
| "clip_improvement": 0.250459 |
| }, |
| "model.layers.43.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954109, |
| "clip_improvement": 0.111494 |
| }, |
| "model.layers.43.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999868, |
| "clip_improvement": 0.250312 |
| }, |
| "model.layers.43.self_attn.k_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 2785280, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.958765, |
| "clip_improvement": 0.111465 |
| }, |
| "model.layers.43.self_attn.o_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955729, |
| "clip_improvement": 0.111737 |
| }, |
| "model.layers.43.self_attn.q_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 12288, |
| 5120 |
| ], |
| "num_rows": 12288, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 33423360, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955155, |
| "clip_improvement": 0.111609 |
| }, |
| "model.layers.43.self_attn.v_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 2785280, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955347, |
| "clip_improvement": 0.11167 |
| }, |
| "model.layers.44.linear_attn.in_proj_a": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.44.linear_attn.in_proj_b": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.44.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 27852800, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955698, |
| "clip_improvement": 0.111134 |
| }, |
| "model.layers.44.linear_attn.in_proj_z": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955713, |
| "clip_improvement": 0.111677 |
| }, |
| "model.layers.44.linear_attn.out_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955037, |
| "clip_improvement": 0.111453 |
| }, |
| "model.layers.44.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999854, |
| "clip_improvement": 0.250403 |
| }, |
| "model.layers.44.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.95438, |
| "clip_improvement": 0.111442 |
| }, |
| "model.layers.44.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999871, |
| "clip_improvement": 0.250432 |
| }, |
| "model.layers.45.linear_attn.in_proj_a": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.45.linear_attn.in_proj_b": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.45.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 27852800, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955984, |
| "clip_improvement": 0.111564 |
| }, |
| "model.layers.45.linear_attn.in_proj_z": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954919, |
| "clip_improvement": 0.111606 |
| }, |
| "model.layers.45.linear_attn.out_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955033, |
| "clip_improvement": 0.111745 |
| }, |
| "model.layers.45.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999836, |
| "clip_improvement": 0.250448 |
| }, |
| "model.layers.45.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954606, |
| "clip_improvement": 0.111444 |
| }, |
| "model.layers.45.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999872, |
| "clip_improvement": 0.25047 |
| }, |
| "model.layers.46.linear_attn.in_proj_a": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.46.linear_attn.in_proj_b": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.46.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 27852800, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955395, |
| "clip_improvement": 0.111414 |
| }, |
| "model.layers.46.linear_attn.in_proj_z": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955245, |
| "clip_improvement": 0.11177 |
| }, |
| "model.layers.46.linear_attn.out_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955282, |
| "clip_improvement": 0.11177 |
| }, |
| "model.layers.46.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999831, |
| "clip_improvement": 0.25042 |
| }, |
| "model.layers.46.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954372, |
| "clip_improvement": 0.111566 |
| }, |
| "model.layers.46.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999831, |
| "clip_improvement": 0.250489 |
| }, |
| "model.layers.47.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999833, |
| "clip_improvement": 0.250788 |
| }, |
| "model.layers.47.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954193, |
| "clip_improvement": 0.111599 |
| }, |
| "model.layers.47.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999865, |
| "clip_improvement": 0.250492 |
| }, |
| "model.layers.47.self_attn.k_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 2785280, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955835, |
| "clip_improvement": 0.111922 |
| }, |
| "model.layers.47.self_attn.o_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.95612, |
| "clip_improvement": 0.111773 |
| }, |
| "model.layers.47.self_attn.q_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 12288, |
| 5120 |
| ], |
| "num_rows": 12288, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 33423360, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955347, |
| "clip_improvement": 0.111727 |
| }, |
| "model.layers.47.self_attn.v_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 2785280, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955396, |
| "clip_improvement": 0.111981 |
| }, |
| "model.layers.48.linear_attn.in_proj_a": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.48.linear_attn.in_proj_b": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.48.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 27852800, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955693, |
| "clip_improvement": 0.111392 |
| }, |
| "model.layers.48.linear_attn.in_proj_z": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955212, |
| "clip_improvement": 0.111621 |
| }, |
| "model.layers.48.linear_attn.out_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955648, |
| "clip_improvement": 0.111403 |
| }, |
| "model.layers.48.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999861, |
| "clip_improvement": 0.25088 |
| }, |
| "model.layers.48.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954881, |
| "clip_improvement": 0.111585 |
| }, |
| "model.layers.48.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999846, |
| "clip_improvement": 0.250546 |
| }, |
| "model.layers.49.linear_attn.in_proj_a": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.49.linear_attn.in_proj_b": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.49.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 27852800, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955832, |
| "clip_improvement": 0.111591 |
| }, |
| "model.layers.49.linear_attn.in_proj_z": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954622, |
| "clip_improvement": 0.111646 |
| }, |
| "model.layers.49.linear_attn.out_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955746, |
| "clip_improvement": 0.111534 |
| }, |
| "model.layers.49.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999845, |
| "clip_improvement": 0.250922 |
| }, |
| "model.layers.49.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954422, |
| "clip_improvement": 0.111833 |
| }, |
| "model.layers.49.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999856, |
| "clip_improvement": 0.250509 |
| }, |
| "model.layers.5.linear_attn.in_proj_a": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.5.linear_attn.in_proj_b": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.5.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 27852800, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954868, |
| "clip_improvement": 0.111796 |
| }, |
| "model.layers.5.linear_attn.in_proj_z": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954374, |
| "clip_improvement": 0.111633 |
| }, |
| "model.layers.5.linear_attn.out_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955383, |
| "clip_improvement": 0.111338 |
| }, |
| "model.layers.5.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999833, |
| "clip_improvement": 0.250193 |
| }, |
| "model.layers.5.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954343, |
| "clip_improvement": 0.111765 |
| }, |
| "model.layers.5.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999852, |
| "clip_improvement": 0.25008 |
| }, |
| "model.layers.50.linear_attn.in_proj_a": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.50.linear_attn.in_proj_b": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.50.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 27852800, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955898, |
| "clip_improvement": 0.111568 |
| }, |
| "model.layers.50.linear_attn.in_proj_z": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955184, |
| "clip_improvement": 0.111782 |
| }, |
| "model.layers.50.linear_attn.out_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955221, |
| "clip_improvement": 0.11127 |
| }, |
| "model.layers.50.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999796, |
| "clip_improvement": 0.252372 |
| }, |
| "model.layers.50.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954739, |
| "clip_improvement": 0.111631 |
| }, |
| "model.layers.50.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999845, |
| "clip_improvement": 0.25089 |
| }, |
| "model.layers.51.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999829, |
| "clip_improvement": 0.251376 |
| }, |
| "model.layers.51.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954689, |
| "clip_improvement": 0.111677 |
| }, |
| "model.layers.51.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999845, |
| "clip_improvement": 0.250395 |
| }, |
| "model.layers.51.self_attn.k_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 2785280, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.956372, |
| "clip_improvement": 0.112017 |
| }, |
| "model.layers.51.self_attn.o_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.956498, |
| "clip_improvement": 0.110943 |
| }, |
| "model.layers.51.self_attn.q_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 12288, |
| 5120 |
| ], |
| "num_rows": 12288, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 33423360, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955149, |
| "clip_improvement": 0.111261 |
| }, |
| "model.layers.51.self_attn.v_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 2785280, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.95835, |
| "clip_improvement": 0.111622 |
| }, |
| "model.layers.52.linear_attn.in_proj_a": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.52.linear_attn.in_proj_b": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.52.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 27852800, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955757, |
| "clip_improvement": 0.111749 |
| }, |
| "model.layers.52.linear_attn.in_proj_z": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955192, |
| "clip_improvement": 0.111994 |
| }, |
| "model.layers.52.linear_attn.out_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955644, |
| "clip_improvement": 0.111805 |
| }, |
| "model.layers.52.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999848, |
| "clip_improvement": 0.250889 |
| }, |
| "model.layers.52.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954677, |
| "clip_improvement": 0.111768 |
| }, |
| "model.layers.52.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999868, |
| "clip_improvement": 0.250569 |
| }, |
| "model.layers.53.linear_attn.in_proj_a": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.53.linear_attn.in_proj_b": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.53.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 27852800, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955112, |
| "clip_improvement": 0.111929 |
| }, |
| "model.layers.53.linear_attn.in_proj_z": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954875, |
| "clip_improvement": 0.111492 |
| }, |
| "model.layers.53.linear_attn.out_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955408, |
| "clip_improvement": 0.111976 |
| }, |
| "model.layers.53.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999828, |
| "clip_improvement": 0.251118 |
| }, |
| "model.layers.53.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954708, |
| "clip_improvement": 0.111873 |
| }, |
| "model.layers.53.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999845, |
| "clip_improvement": 0.250399 |
| }, |
| "model.layers.54.linear_attn.in_proj_a": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.54.linear_attn.in_proj_b": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.54.linear_attn.in_proj_qkv": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 21299200, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999778, |
| "clip_improvement": 0.250885 |
| }, |
| "model.layers.54.linear_attn.in_proj_z": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954549, |
| "clip_improvement": 0.111832 |
| }, |
| "model.layers.54.linear_attn.out_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.95529, |
| "clip_improvement": 0.112607 |
| }, |
| "model.layers.54.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999819, |
| "clip_improvement": 0.25104 |
| }, |
| "model.layers.54.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955171, |
| "clip_improvement": 0.111786 |
| }, |
| "model.layers.54.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999835, |
| "clip_improvement": 0.250397 |
| }, |
| "model.layers.55.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999864, |
| "clip_improvement": 0.250581 |
| }, |
| "model.layers.55.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954804, |
| "clip_improvement": 0.111901 |
| }, |
| "model.layers.55.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999802, |
| "clip_improvement": 0.250183 |
| }, |
| "model.layers.55.self_attn.k_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 2785280, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.959204, |
| "clip_improvement": 0.111035 |
| }, |
| "model.layers.55.self_attn.o_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 12779520, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999801, |
| "clip_improvement": 0.252846 |
| }, |
| "model.layers.55.self_attn.q_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 12288, |
| 5120 |
| ], |
| "num_rows": 12288, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 25559040, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999866, |
| "clip_improvement": 0.250785 |
| }, |
| "model.layers.55.self_attn.v_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 2785280, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.957275, |
| "clip_improvement": 0.110343 |
| }, |
| "model.layers.56.linear_attn.in_proj_a": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.56.linear_attn.in_proj_b": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.56.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 27852800, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955205, |
| "clip_improvement": 0.112076 |
| }, |
| "model.layers.56.linear_attn.in_proj_z": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954736, |
| "clip_improvement": 0.111756 |
| }, |
| "model.layers.56.linear_attn.out_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955355, |
| "clip_improvement": 0.112062 |
| }, |
| "model.layers.56.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999868, |
| "clip_improvement": 0.250452 |
| }, |
| "model.layers.56.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954389, |
| "clip_improvement": 0.111774 |
| }, |
| "model.layers.56.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999841, |
| "clip_improvement": 0.250318 |
| }, |
| "model.layers.57.linear_attn.in_proj_a": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.57.linear_attn.in_proj_b": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.57.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 27852800, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955303, |
| "clip_improvement": 0.112053 |
| }, |
| "model.layers.57.linear_attn.in_proj_z": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954313, |
| "clip_improvement": 0.111759 |
| }, |
| "model.layers.57.linear_attn.out_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955802, |
| "clip_improvement": 0.112172 |
| }, |
| "model.layers.57.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999841, |
| "clip_improvement": 0.250379 |
| }, |
| "model.layers.57.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954439, |
| "clip_improvement": 0.111959 |
| }, |
| "model.layers.57.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999874, |
| "clip_improvement": 0.250097 |
| }, |
| "model.layers.58.linear_attn.in_proj_a": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.58.linear_attn.in_proj_b": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.58.linear_attn.in_proj_qkv": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 21299200, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999814, |
| "clip_improvement": 0.250903 |
| }, |
| "model.layers.58.linear_attn.in_proj_z": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 12779520, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.99989, |
| "clip_improvement": 0.250319 |
| }, |
| "model.layers.58.linear_attn.out_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 12779520, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999829, |
| "clip_improvement": 0.25211 |
| }, |
| "model.layers.58.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.99981, |
| "clip_improvement": 0.250733 |
| }, |
| "model.layers.58.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954584, |
| "clip_improvement": 0.11205 |
| }, |
| "model.layers.58.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999856, |
| "clip_improvement": 0.250113 |
| }, |
| "model.layers.59.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999864, |
| "clip_improvement": 0.250645 |
| }, |
| "model.layers.59.mlp.gate_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999848, |
| "clip_improvement": 0.25073 |
| }, |
| "model.layers.59.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999821, |
| "clip_improvement": 0.250441 |
| }, |
| "model.layers.59.self_attn.k_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 2785280, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.959399, |
| "clip_improvement": 0.110837 |
| }, |
| "model.layers.59.self_attn.o_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955957, |
| "clip_improvement": 0.111146 |
| }, |
| "model.layers.59.self_attn.q_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 12288, |
| 5120 |
| ], |
| "num_rows": 12288, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 33423360, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954901, |
| "clip_improvement": 0.111785 |
| }, |
| "model.layers.59.self_attn.v_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 2785280, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.957861, |
| "clip_improvement": 0.113004 |
| }, |
| "model.layers.6.linear_attn.in_proj_a": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.6.linear_attn.in_proj_b": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.6.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 27852800, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955254, |
| "clip_improvement": 0.1117 |
| }, |
| "model.layers.6.linear_attn.in_proj_z": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.95518, |
| "clip_improvement": 0.111621 |
| }, |
| "model.layers.6.linear_attn.out_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955542, |
| "clip_improvement": 0.112362 |
| }, |
| "model.layers.6.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999703, |
| "clip_improvement": 0.252978 |
| }, |
| "model.layers.6.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.953823, |
| "clip_improvement": 0.111815 |
| }, |
| "model.layers.6.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999835, |
| "clip_improvement": 0.250559 |
| }, |
| "model.layers.60.linear_attn.in_proj_a": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.60.linear_attn.in_proj_b": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.60.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 27852800, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955249, |
| "clip_improvement": 0.112038 |
| }, |
| "model.layers.60.linear_attn.in_proj_z": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954533, |
| "clip_improvement": 0.111869 |
| }, |
| "model.layers.60.linear_attn.out_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955074, |
| "clip_improvement": 0.111918 |
| }, |
| "model.layers.60.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999843, |
| "clip_improvement": 0.250864 |
| }, |
| "model.layers.60.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.95441, |
| "clip_improvement": 0.111778 |
| }, |
| "model.layers.60.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999849, |
| "clip_improvement": 0.250258 |
| }, |
| "model.layers.61.linear_attn.in_proj_a": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.61.linear_attn.in_proj_b": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.61.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 27852800, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955271, |
| "clip_improvement": 0.111855 |
| }, |
| "model.layers.61.linear_attn.in_proj_z": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955009, |
| "clip_improvement": 0.112015 |
| }, |
| "model.layers.61.linear_attn.out_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955607, |
| "clip_improvement": 0.111875 |
| }, |
| "model.layers.61.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999843, |
| "clip_improvement": 0.250382 |
| }, |
| "model.layers.61.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954732, |
| "clip_improvement": 0.111989 |
| }, |
| "model.layers.61.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999822, |
| "clip_improvement": 0.250265 |
| }, |
| "model.layers.62.linear_attn.in_proj_a": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.62.linear_attn.in_proj_b": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.62.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 27852800, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955854, |
| "clip_improvement": 0.111737 |
| }, |
| "model.layers.62.linear_attn.in_proj_z": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954631, |
| "clip_improvement": 0.112326 |
| }, |
| "model.layers.62.linear_attn.out_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.956116, |
| "clip_improvement": 0.111719 |
| }, |
| "model.layers.62.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999843, |
| "clip_improvement": 0.251275 |
| }, |
| "model.layers.62.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954667, |
| "clip_improvement": 0.111817 |
| }, |
| "model.layers.62.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999818, |
| "clip_improvement": 0.250395 |
| }, |
| "model.layers.63.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999875, |
| "clip_improvement": 0.252275 |
| }, |
| "model.layers.63.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954699, |
| "clip_improvement": 0.111917 |
| }, |
| "model.layers.63.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999833, |
| "clip_improvement": 0.250269 |
| }, |
| "model.layers.63.self_attn.k_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 2785280, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.959839, |
| "clip_improvement": 0.111721 |
| }, |
| "model.layers.63.self_attn.o_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.959017, |
| "clip_improvement": 0.110433 |
| }, |
| "model.layers.63.self_attn.q_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 12288, |
| 5120 |
| ], |
| "num_rows": 12288, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 33423360, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955326, |
| "clip_improvement": 0.111927 |
| }, |
| "model.layers.63.self_attn.v_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 2785280, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955786, |
| "clip_improvement": 0.112904 |
| }, |
| "model.layers.7.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999831, |
| "clip_improvement": 0.250414 |
| }, |
| "model.layers.7.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954307, |
| "clip_improvement": 0.111636 |
| }, |
| "model.layers.7.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999843, |
| "clip_improvement": 0.250154 |
| }, |
| "model.layers.7.self_attn.k_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 2785280, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.958765, |
| "clip_improvement": 0.112238 |
| }, |
| "model.layers.7.self_attn.o_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954407, |
| "clip_improvement": 0.111939 |
| }, |
| "model.layers.7.self_attn.q_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 12288, |
| 5120 |
| ], |
| "num_rows": 12288, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 33423360, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955052, |
| "clip_improvement": 0.111794 |
| }, |
| "model.layers.7.self_attn.v_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 2785280, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.956201, |
| "clip_improvement": 0.112123 |
| }, |
| "model.layers.8.linear_attn.in_proj_a": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.8.linear_attn.in_proj_b": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.8.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 27852800, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955012, |
| "clip_improvement": 0.111602 |
| }, |
| "model.layers.8.linear_attn.in_proj_z": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954907, |
| "clip_improvement": 0.111852 |
| }, |
| "model.layers.8.linear_attn.out_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954948, |
| "clip_improvement": 0.111565 |
| }, |
| "model.layers.8.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999869, |
| "clip_improvement": 0.250149 |
| }, |
| "model.layers.8.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.953932, |
| "clip_improvement": 0.111709 |
| }, |
| "model.layers.8.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999854, |
| "clip_improvement": 0.249985 |
| }, |
| "model.layers.9.linear_attn.in_proj_a": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.9.linear_attn.in_proj_b": { |
| "bits": 8, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 48, |
| 5120 |
| ], |
| "num_rows": 48, |
| "num_groups": 40, |
| "words_per_group": 32, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 253440, |
| "bits_per_weight": 8.25, |
| "clipped_fraction": 0.0, |
| "clip_improvement": 0.0 |
| }, |
| "model.layers.9.linear_attn.in_proj_qkv": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 27852800, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955198, |
| "clip_improvement": 0.112071 |
| }, |
| "model.layers.9.linear_attn.in_proj_z": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.955151, |
| "clip_improvement": 0.112055 |
| }, |
| "model.layers.9.linear_attn.out_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 16711680, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.954606, |
| "clip_improvement": 0.111772 |
| }, |
| "model.layers.9.mlp.down_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999829, |
| "clip_improvement": 0.250178 |
| }, |
| "model.layers.9.mlp.gate_proj": { |
| "bits": 4, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 16, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 47349760, |
| "bits_per_weight": 4.25, |
| "clipped_fraction": 0.953882, |
| "clip_improvement": 0.111555 |
| }, |
| "model.layers.9.mlp.up_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 17408, |
| 5120 |
| ], |
| "num_rows": 17408, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 36208640, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999842, |
| "clip_improvement": 0.250002 |
| } |
| } |
| } |
