| { |
| "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": 3, |
| "out_features": 248320 |
| }, |
| "model.embed_tokens": { |
| "bits": 3, |
| "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": 3, |
| "out_features": 10240 |
| }, |
| "model.layers.0.linear_attn.in_proj_z": { |
| "bits": 3, |
| "out_features": 6144 |
| }, |
| "model.layers.0.linear_attn.out_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.0.mlp.down_proj": { |
| "bits": 2, |
| "out_features": 5120 |
| }, |
| "model.layers.0.mlp.gate_proj": { |
| "bits": 2, |
| "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": 3, |
| "out_features": 10240 |
| }, |
| "model.layers.1.linear_attn.in_proj_z": { |
| "bits": 3, |
| "out_features": 6144 |
| }, |
| "model.layers.1.linear_attn.out_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.1.mlp.down_proj": { |
| "bits": 2, |
| "out_features": 5120 |
| }, |
| "model.layers.1.mlp.gate_proj": { |
| "bits": 2, |
| "out_features": 17408 |
| }, |
| "model.layers.1.mlp.up_proj": { |
| "bits": 2, |
| "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": 3, |
| "out_features": 10240 |
| }, |
| "model.layers.10.linear_attn.in_proj_z": { |
| "bits": 3, |
| "out_features": 6144 |
| }, |
| "model.layers.10.linear_attn.out_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.10.mlp.down_proj": { |
| "bits": 2, |
| "out_features": 5120 |
| }, |
| "model.layers.10.mlp.gate_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.10.mlp.up_proj": { |
| "bits": 2, |
| "out_features": 17408 |
| }, |
| "model.layers.11.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.11.mlp.gate_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.11.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.11.self_attn.k_proj": { |
| "bits": 3, |
| "out_features": 1024 |
| }, |
| "model.layers.11.self_attn.o_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.11.self_attn.q_proj": { |
| "bits": 3, |
| "out_features": 12288 |
| }, |
| "model.layers.11.self_attn.v_proj": { |
| "bits": 3, |
| "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": 3, |
| "out_features": 10240 |
| }, |
| "model.layers.12.linear_attn.in_proj_z": { |
| "bits": 3, |
| "out_features": 6144 |
| }, |
| "model.layers.12.linear_attn.out_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.12.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.12.mlp.gate_proj": { |
| "bits": 3, |
| "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": 3, |
| "out_features": 10240 |
| }, |
| "model.layers.13.linear_attn.in_proj_z": { |
| "bits": 3, |
| "out_features": 6144 |
| }, |
| "model.layers.13.linear_attn.out_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.13.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.13.mlp.gate_proj": { |
| "bits": 3, |
| "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": 3, |
| "out_features": 10240 |
| }, |
| "model.layers.14.linear_attn.in_proj_z": { |
| "bits": 3, |
| "out_features": 6144 |
| }, |
| "model.layers.14.linear_attn.out_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.14.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.14.mlp.gate_proj": { |
| "bits": 3, |
| "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": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.15.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.15.self_attn.k_proj": { |
| "bits": 3, |
| "out_features": 1024 |
| }, |
| "model.layers.15.self_attn.o_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.15.self_attn.q_proj": { |
| "bits": 3, |
| "out_features": 12288 |
| }, |
| "model.layers.15.self_attn.v_proj": { |
| "bits": 3, |
| "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": 3, |
| "out_features": 10240 |
| }, |
| "model.layers.16.linear_attn.in_proj_z": { |
| "bits": 3, |
| "out_features": 6144 |
| }, |
| "model.layers.16.linear_attn.out_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.16.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.16.mlp.gate_proj": { |
| "bits": 3, |
| "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": 3, |
| "out_features": 10240 |
| }, |
| "model.layers.17.linear_attn.in_proj_z": { |
| "bits": 3, |
| "out_features": 6144 |
| }, |
| "model.layers.17.linear_attn.out_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.17.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.17.mlp.gate_proj": { |
| "bits": 3, |
| "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": 3, |
| "out_features": 10240 |
| }, |
| "model.layers.18.linear_attn.in_proj_z": { |
| "bits": 3, |
| "out_features": 6144 |
| }, |
| "model.layers.18.linear_attn.out_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.18.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.18.mlp.gate_proj": { |
| "bits": 3, |
| "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": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.19.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.19.self_attn.k_proj": { |
| "bits": 3, |
| "out_features": 1024 |
| }, |
| "model.layers.19.self_attn.o_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.19.self_attn.q_proj": { |
| "bits": 3, |
| "out_features": 12288 |
| }, |
| "model.layers.19.self_attn.v_proj": { |
| "bits": 3, |
| "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": 3, |
| "out_features": 10240 |
| }, |
| "model.layers.2.linear_attn.in_proj_z": { |
| "bits": 3, |
| "out_features": 6144 |
| }, |
| "model.layers.2.linear_attn.out_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.2.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.2.mlp.gate_proj": { |
| "bits": 2, |
| "out_features": 17408 |
| }, |
| "model.layers.2.mlp.up_proj": { |
| "bits": 2, |
| "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": 3, |
| "out_features": 10240 |
| }, |
| "model.layers.20.linear_attn.in_proj_z": { |
| "bits": 3, |
| "out_features": 6144 |
| }, |
| "model.layers.20.linear_attn.out_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.20.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.20.mlp.gate_proj": { |
| "bits": 3, |
| "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": 3, |
| "out_features": 10240 |
| }, |
| "model.layers.21.linear_attn.in_proj_z": { |
| "bits": 3, |
| "out_features": 6144 |
| }, |
| "model.layers.21.linear_attn.out_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.21.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.21.mlp.gate_proj": { |
| "bits": 3, |
| "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": 3, |
| "out_features": 10240 |
| }, |
| "model.layers.22.linear_attn.in_proj_z": { |
| "bits": 3, |
| "out_features": 6144 |
| }, |
| "model.layers.22.linear_attn.out_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.22.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.22.mlp.gate_proj": { |
| "bits": 3, |
| "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": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.23.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.23.self_attn.k_proj": { |
| "bits": 3, |
| "out_features": 1024 |
| }, |
| "model.layers.23.self_attn.o_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.23.self_attn.q_proj": { |
| "bits": 3, |
| "out_features": 12288 |
| }, |
| "model.layers.23.self_attn.v_proj": { |
| "bits": 3, |
| "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": 3, |
| "out_features": 10240 |
| }, |
| "model.layers.24.linear_attn.in_proj_z": { |
| "bits": 3, |
| "out_features": 6144 |
| }, |
| "model.layers.24.linear_attn.out_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.24.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.24.mlp.gate_proj": { |
| "bits": 3, |
| "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": 3, |
| "out_features": 10240 |
| }, |
| "model.layers.25.linear_attn.in_proj_z": { |
| "bits": 3, |
| "out_features": 6144 |
| }, |
| "model.layers.25.linear_attn.out_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.25.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.25.mlp.gate_proj": { |
| "bits": 3, |
| "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": 3, |
| "out_features": 10240 |
| }, |
| "model.layers.26.linear_attn.in_proj_z": { |
| "bits": 3, |
| "out_features": 6144 |
| }, |
| "model.layers.26.linear_attn.out_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.26.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.26.mlp.gate_proj": { |
| "bits": 3, |
| "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": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.27.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.27.self_attn.k_proj": { |
| "bits": 3, |
| "out_features": 1024 |
| }, |
| "model.layers.27.self_attn.o_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.27.self_attn.q_proj": { |
| "bits": 3, |
| "out_features": 12288 |
| }, |
| "model.layers.27.self_attn.v_proj": { |
| "bits": 3, |
| "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": 3, |
| "out_features": 10240 |
| }, |
| "model.layers.28.linear_attn.in_proj_z": { |
| "bits": 3, |
| "out_features": 6144 |
| }, |
| "model.layers.28.linear_attn.out_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.28.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.28.mlp.gate_proj": { |
| "bits": 3, |
| "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": 3, |
| "out_features": 10240 |
| }, |
| "model.layers.29.linear_attn.in_proj_z": { |
| "bits": 3, |
| "out_features": 6144 |
| }, |
| "model.layers.29.linear_attn.out_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.29.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.29.mlp.gate_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.29.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.3.mlp.down_proj": { |
| "bits": 2, |
| "out_features": 5120 |
| }, |
| "model.layers.3.mlp.gate_proj": { |
| "bits": 2, |
| "out_features": 17408 |
| }, |
| "model.layers.3.mlp.up_proj": { |
| "bits": 2, |
| "out_features": 17408 |
| }, |
| "model.layers.3.self_attn.k_proj": { |
| "bits": 3, |
| "out_features": 1024 |
| }, |
| "model.layers.3.self_attn.o_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.3.self_attn.q_proj": { |
| "bits": 3, |
| "out_features": 12288 |
| }, |
| "model.layers.3.self_attn.v_proj": { |
| "bits": 3, |
| "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": 3, |
| "out_features": 10240 |
| }, |
| "model.layers.30.linear_attn.in_proj_z": { |
| "bits": 3, |
| "out_features": 6144 |
| }, |
| "model.layers.30.linear_attn.out_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.30.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.30.mlp.gate_proj": { |
| "bits": 3, |
| "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": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.31.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.31.self_attn.k_proj": { |
| "bits": 3, |
| "out_features": 1024 |
| }, |
| "model.layers.31.self_attn.o_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.31.self_attn.q_proj": { |
| "bits": 3, |
| "out_features": 12288 |
| }, |
| "model.layers.31.self_attn.v_proj": { |
| "bits": 3, |
| "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": 3, |
| "out_features": 10240 |
| }, |
| "model.layers.32.linear_attn.in_proj_z": { |
| "bits": 3, |
| "out_features": 6144 |
| }, |
| "model.layers.32.linear_attn.out_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.32.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.32.mlp.gate_proj": { |
| "bits": 3, |
| "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": 3, |
| "out_features": 10240 |
| }, |
| "model.layers.33.linear_attn.in_proj_z": { |
| "bits": 3, |
| "out_features": 6144 |
| }, |
| "model.layers.33.linear_attn.out_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.33.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.33.mlp.gate_proj": { |
| "bits": 3, |
| "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": 3, |
| "out_features": 10240 |
| }, |
| "model.layers.34.linear_attn.in_proj_z": { |
| "bits": 3, |
| "out_features": 6144 |
| }, |
| "model.layers.34.linear_attn.out_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.34.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.34.mlp.gate_proj": { |
| "bits": 3, |
| "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": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.35.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.35.self_attn.k_proj": { |
| "bits": 3, |
| "out_features": 1024 |
| }, |
| "model.layers.35.self_attn.o_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.35.self_attn.q_proj": { |
| "bits": 3, |
| "out_features": 12288 |
| }, |
| "model.layers.35.self_attn.v_proj": { |
| "bits": 3, |
| "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": 3, |
| "out_features": 10240 |
| }, |
| "model.layers.36.linear_attn.in_proj_z": { |
| "bits": 3, |
| "out_features": 6144 |
| }, |
| "model.layers.36.linear_attn.out_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.36.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.36.mlp.gate_proj": { |
| "bits": 3, |
| "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": 3, |
| "out_features": 10240 |
| }, |
| "model.layers.37.linear_attn.in_proj_z": { |
| "bits": 3, |
| "out_features": 6144 |
| }, |
| "model.layers.37.linear_attn.out_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.37.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.37.mlp.gate_proj": { |
| "bits": 3, |
| "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": 3, |
| "out_features": 10240 |
| }, |
| "model.layers.38.linear_attn.in_proj_z": { |
| "bits": 3, |
| "out_features": 6144 |
| }, |
| "model.layers.38.linear_attn.out_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.38.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.38.mlp.gate_proj": { |
| "bits": 3, |
| "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": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.39.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.39.self_attn.k_proj": { |
| "bits": 3, |
| "out_features": 1024 |
| }, |
| "model.layers.39.self_attn.o_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.39.self_attn.q_proj": { |
| "bits": 3, |
| "out_features": 12288 |
| }, |
| "model.layers.39.self_attn.v_proj": { |
| "bits": 3, |
| "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": 2, |
| "out_features": 10240 |
| }, |
| "model.layers.4.linear_attn.in_proj_z": { |
| "bits": 2, |
| "out_features": 6144 |
| }, |
| "model.layers.4.linear_attn.out_proj": { |
| "bits": 2, |
| "out_features": 5120 |
| }, |
| "model.layers.4.mlp.down_proj": { |
| "bits": 2, |
| "out_features": 5120 |
| }, |
| "model.layers.4.mlp.gate_proj": { |
| "bits": 2, |
| "out_features": 17408 |
| }, |
| "model.layers.4.mlp.up_proj": { |
| "bits": 2, |
| "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": 3, |
| "out_features": 10240 |
| }, |
| "model.layers.40.linear_attn.in_proj_z": { |
| "bits": 3, |
| "out_features": 6144 |
| }, |
| "model.layers.40.linear_attn.out_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.40.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.40.mlp.gate_proj": { |
| "bits": 3, |
| "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": 3, |
| "out_features": 10240 |
| }, |
| "model.layers.41.linear_attn.in_proj_z": { |
| "bits": 3, |
| "out_features": 6144 |
| }, |
| "model.layers.41.linear_attn.out_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.41.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.41.mlp.gate_proj": { |
| "bits": 3, |
| "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": 3, |
| "out_features": 10240 |
| }, |
| "model.layers.42.linear_attn.in_proj_z": { |
| "bits": 3, |
| "out_features": 6144 |
| }, |
| "model.layers.42.linear_attn.out_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.42.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.42.mlp.gate_proj": { |
| "bits": 3, |
| "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": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.43.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.43.self_attn.k_proj": { |
| "bits": 3, |
| "out_features": 1024 |
| }, |
| "model.layers.43.self_attn.o_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.43.self_attn.q_proj": { |
| "bits": 3, |
| "out_features": 12288 |
| }, |
| "model.layers.43.self_attn.v_proj": { |
| "bits": 3, |
| "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": 3, |
| "out_features": 10240 |
| }, |
| "model.layers.44.linear_attn.in_proj_z": { |
| "bits": 3, |
| "out_features": 6144 |
| }, |
| "model.layers.44.linear_attn.out_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.44.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.44.mlp.gate_proj": { |
| "bits": 3, |
| "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": 3, |
| "out_features": 10240 |
| }, |
| "model.layers.45.linear_attn.in_proj_z": { |
| "bits": 3, |
| "out_features": 6144 |
| }, |
| "model.layers.45.linear_attn.out_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.45.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.45.mlp.gate_proj": { |
| "bits": 3, |
| "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": 2, |
| "out_features": 10240 |
| }, |
| "model.layers.46.linear_attn.in_proj_z": { |
| "bits": 3, |
| "out_features": 6144 |
| }, |
| "model.layers.46.linear_attn.out_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.46.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.46.mlp.gate_proj": { |
| "bits": 3, |
| "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": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.47.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.47.self_attn.k_proj": { |
| "bits": 2, |
| "out_features": 1024 |
| }, |
| "model.layers.47.self_attn.o_proj": { |
| "bits": 2, |
| "out_features": 5120 |
| }, |
| "model.layers.47.self_attn.q_proj": { |
| "bits": 2, |
| "out_features": 12288 |
| }, |
| "model.layers.47.self_attn.v_proj": { |
| "bits": 3, |
| "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": 2, |
| "out_features": 10240 |
| }, |
| "model.layers.48.linear_attn.in_proj_z": { |
| "bits": 2, |
| "out_features": 6144 |
| }, |
| "model.layers.48.linear_attn.out_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.48.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.48.mlp.gate_proj": { |
| "bits": 3, |
| "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": 2, |
| "out_features": 10240 |
| }, |
| "model.layers.49.linear_attn.in_proj_z": { |
| "bits": 2, |
| "out_features": 6144 |
| }, |
| "model.layers.49.linear_attn.out_proj": { |
| "bits": 2, |
| "out_features": 5120 |
| }, |
| "model.layers.49.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.49.mlp.gate_proj": { |
| "bits": 3, |
| "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": 3, |
| "out_features": 10240 |
| }, |
| "model.layers.5.linear_attn.in_proj_z": { |
| "bits": 3, |
| "out_features": 6144 |
| }, |
| "model.layers.5.linear_attn.out_proj": { |
| "bits": 2, |
| "out_features": 5120 |
| }, |
| "model.layers.5.mlp.down_proj": { |
| "bits": 2, |
| "out_features": 5120 |
| }, |
| "model.layers.5.mlp.gate_proj": { |
| "bits": 2, |
| "out_features": 17408 |
| }, |
| "model.layers.5.mlp.up_proj": { |
| "bits": 2, |
| "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": 2, |
| "out_features": 10240 |
| }, |
| "model.layers.50.linear_attn.in_proj_z": { |
| "bits": 2, |
| "out_features": 6144 |
| }, |
| "model.layers.50.linear_attn.out_proj": { |
| "bits": 2, |
| "out_features": 5120 |
| }, |
| "model.layers.50.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.50.mlp.gate_proj": { |
| "bits": 3, |
| "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": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.51.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.51.self_attn.k_proj": { |
| "bits": 2, |
| "out_features": 1024 |
| }, |
| "model.layers.51.self_attn.o_proj": { |
| "bits": 2, |
| "out_features": 5120 |
| }, |
| "model.layers.51.self_attn.q_proj": { |
| "bits": 2, |
| "out_features": 12288 |
| }, |
| "model.layers.51.self_attn.v_proj": { |
| "bits": 2, |
| "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": 2, |
| "out_features": 10240 |
| }, |
| "model.layers.52.linear_attn.in_proj_z": { |
| "bits": 2, |
| "out_features": 6144 |
| }, |
| "model.layers.52.linear_attn.out_proj": { |
| "bits": 2, |
| "out_features": 5120 |
| }, |
| "model.layers.52.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.52.mlp.gate_proj": { |
| "bits": 3, |
| "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": 2, |
| "out_features": 10240 |
| }, |
| "model.layers.53.linear_attn.in_proj_z": { |
| "bits": 2, |
| "out_features": 6144 |
| }, |
| "model.layers.53.linear_attn.out_proj": { |
| "bits": 2, |
| "out_features": 5120 |
| }, |
| "model.layers.53.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.53.mlp.gate_proj": { |
| "bits": 3, |
| "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": 2, |
| "out_features": 10240 |
| }, |
| "model.layers.54.linear_attn.in_proj_z": { |
| "bits": 2, |
| "out_features": 6144 |
| }, |
| "model.layers.54.linear_attn.out_proj": { |
| "bits": 2, |
| "out_features": 5120 |
| }, |
| "model.layers.54.mlp.down_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.54.mlp.gate_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.54.mlp.up_proj": { |
| "bits": 2, |
| "out_features": 17408 |
| }, |
| "model.layers.55.mlp.down_proj": { |
| "bits": 2, |
| "out_features": 5120 |
| }, |
| "model.layers.55.mlp.gate_proj": { |
| "bits": 2, |
| "out_features": 17408 |
| }, |
| "model.layers.55.mlp.up_proj": { |
| "bits": 2, |
| "out_features": 17408 |
| }, |
| "model.layers.55.self_attn.k_proj": { |
| "bits": 2, |
| "out_features": 1024 |
| }, |
| "model.layers.55.self_attn.o_proj": { |
| "bits": 2, |
| "out_features": 5120 |
| }, |
| "model.layers.55.self_attn.q_proj": { |
| "bits": 2, |
| "out_features": 12288 |
| }, |
| "model.layers.55.self_attn.v_proj": { |
| "bits": 2, |
| "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": 2, |
| "out_features": 10240 |
| }, |
| "model.layers.56.linear_attn.in_proj_z": { |
| "bits": 2, |
| "out_features": 6144 |
| }, |
| "model.layers.56.linear_attn.out_proj": { |
| "bits": 2, |
| "out_features": 5120 |
| }, |
| "model.layers.56.mlp.down_proj": { |
| "bits": 2, |
| "out_features": 5120 |
| }, |
| "model.layers.56.mlp.gate_proj": { |
| "bits": 2, |
| "out_features": 17408 |
| }, |
| "model.layers.56.mlp.up_proj": { |
| "bits": 2, |
| "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": 2, |
| "out_features": 10240 |
| }, |
| "model.layers.57.linear_attn.in_proj_z": { |
| "bits": 2, |
| "out_features": 6144 |
| }, |
| "model.layers.57.linear_attn.out_proj": { |
| "bits": 2, |
| "out_features": 5120 |
| }, |
| "model.layers.57.mlp.down_proj": { |
| "bits": 2, |
| "out_features": 5120 |
| }, |
| "model.layers.57.mlp.gate_proj": { |
| "bits": 2, |
| "out_features": 17408 |
| }, |
| "model.layers.57.mlp.up_proj": { |
| "bits": 2, |
| "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": 2, |
| "out_features": 10240 |
| }, |
| "model.layers.58.linear_attn.in_proj_z": { |
| "bits": 2, |
| "out_features": 6144 |
| }, |
| "model.layers.58.linear_attn.out_proj": { |
| "bits": 2, |
| "out_features": 5120 |
| }, |
| "model.layers.58.mlp.down_proj": { |
| "bits": 2, |
| "out_features": 5120 |
| }, |
| "model.layers.58.mlp.gate_proj": { |
| "bits": 2, |
| "out_features": 17408 |
| }, |
| "model.layers.58.mlp.up_proj": { |
| "bits": 2, |
| "out_features": 17408 |
| }, |
| "model.layers.59.mlp.down_proj": { |
| "bits": 2, |
| "out_features": 5120 |
| }, |
| "model.layers.59.mlp.gate_proj": { |
| "bits": 2, |
| "out_features": 17408 |
| }, |
| "model.layers.59.mlp.up_proj": { |
| "bits": 2, |
| "out_features": 17408 |
| }, |
| "model.layers.59.self_attn.k_proj": { |
| "bits": 2, |
| "out_features": 1024 |
| }, |
| "model.layers.59.self_attn.o_proj": { |
| "bits": 2, |
| "out_features": 5120 |
| }, |
| "model.layers.59.self_attn.q_proj": { |
| "bits": 2, |
| "out_features": 12288 |
| }, |
| "model.layers.59.self_attn.v_proj": { |
| "bits": 2, |
| "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": 3, |
| "out_features": 10240 |
| }, |
| "model.layers.6.linear_attn.in_proj_z": { |
| "bits": 3, |
| "out_features": 6144 |
| }, |
| "model.layers.6.linear_attn.out_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.6.mlp.down_proj": { |
| "bits": 2, |
| "out_features": 5120 |
| }, |
| "model.layers.6.mlp.gate_proj": { |
| "bits": 3, |
| "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": 2, |
| "out_features": 10240 |
| }, |
| "model.layers.60.linear_attn.in_proj_z": { |
| "bits": 2, |
| "out_features": 6144 |
| }, |
| "model.layers.60.linear_attn.out_proj": { |
| "bits": 2, |
| "out_features": 5120 |
| }, |
| "model.layers.60.mlp.down_proj": { |
| "bits": 2, |
| "out_features": 5120 |
| }, |
| "model.layers.60.mlp.gate_proj": { |
| "bits": 2, |
| "out_features": 17408 |
| }, |
| "model.layers.60.mlp.up_proj": { |
| "bits": 2, |
| "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": 2, |
| "out_features": 10240 |
| }, |
| "model.layers.61.linear_attn.in_proj_z": { |
| "bits": 2, |
| "out_features": 6144 |
| }, |
| "model.layers.61.linear_attn.out_proj": { |
| "bits": 2, |
| "out_features": 5120 |
| }, |
| "model.layers.61.mlp.down_proj": { |
| "bits": 2, |
| "out_features": 5120 |
| }, |
| "model.layers.61.mlp.gate_proj": { |
| "bits": 2, |
| "out_features": 17408 |
| }, |
| "model.layers.61.mlp.up_proj": { |
| "bits": 2, |
| "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": 2, |
| "out_features": 10240 |
| }, |
| "model.layers.62.linear_attn.in_proj_z": { |
| "bits": 2, |
| "out_features": 6144 |
| }, |
| "model.layers.62.linear_attn.out_proj": { |
| "bits": 2, |
| "out_features": 5120 |
| }, |
| "model.layers.62.mlp.down_proj": { |
| "bits": 2, |
| "out_features": 5120 |
| }, |
| "model.layers.62.mlp.gate_proj": { |
| "bits": 2, |
| "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": 2, |
| "out_features": 17408 |
| }, |
| "model.layers.63.mlp.up_proj": { |
| "bits": 3, |
| "out_features": 17408 |
| }, |
| "model.layers.63.self_attn.k_proj": { |
| "bits": 3, |
| "out_features": 1024 |
| }, |
| "model.layers.63.self_attn.o_proj": { |
| "bits": 2, |
| "out_features": 5120 |
| }, |
| "model.layers.63.self_attn.q_proj": { |
| "bits": 2, |
| "out_features": 12288 |
| }, |
| "model.layers.63.self_attn.v_proj": { |
| "bits": 2, |
| "out_features": 1024 |
| }, |
| "model.layers.7.mlp.down_proj": { |
| "bits": 2, |
| "out_features": 5120 |
| }, |
| "model.layers.7.mlp.gate_proj": { |
| "bits": 2, |
| "out_features": 17408 |
| }, |
| "model.layers.7.mlp.up_proj": { |
| "bits": 2, |
| "out_features": 17408 |
| }, |
| "model.layers.7.self_attn.k_proj": { |
| "bits": 3, |
| "out_features": 1024 |
| }, |
| "model.layers.7.self_attn.o_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.7.self_attn.q_proj": { |
| "bits": 3, |
| "out_features": 12288 |
| }, |
| "model.layers.7.self_attn.v_proj": { |
| "bits": 3, |
| "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": 3, |
| "out_features": 10240 |
| }, |
| "model.layers.8.linear_attn.in_proj_z": { |
| "bits": 3, |
| "out_features": 6144 |
| }, |
| "model.layers.8.linear_attn.out_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.8.mlp.down_proj": { |
| "bits": 2, |
| "out_features": 5120 |
| }, |
| "model.layers.8.mlp.gate_proj": { |
| "bits": 2, |
| "out_features": 17408 |
| }, |
| "model.layers.8.mlp.up_proj": { |
| "bits": 2, |
| "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": 3, |
| "out_features": 10240 |
| }, |
| "model.layers.9.linear_attn.in_proj_z": { |
| "bits": 3, |
| "out_features": 6144 |
| }, |
| "model.layers.9.linear_attn.out_proj": { |
| "bits": 3, |
| "out_features": 5120 |
| }, |
| "model.layers.9.mlp.down_proj": { |
| "bits": 2, |
| "out_features": 5120 |
| }, |
| "model.layers.9.mlp.gate_proj": { |
| "bits": 2, |
| "out_features": 17408 |
| }, |
| "model.layers.9.mlp.up_proj": { |
| "bits": 2, |
| "out_features": 17408 |
| } |
| } |
| }, |
| "files": [ |
| "model-00001-of-00003.safetensors", |
| "model-00002-of-00003.safetensors", |
| "model-00003-of-00003.safetensors", |
| "model.safetensors.index.json" |
| ], |
| "accounting": { |
| "packed_nbytes": 10080337920, |
| "dense_nbytes": 5291008, |
| "total_nbytes": 10085628928, |
| "quantized_elements": 26893352960, |
| "average_bits": 2.998611 |
| }, |
| "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": "3.00 avg bits", |
| "predicted_nbytes": 10085628928 |
| }, |
| "layers": { |
| "lm_head": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 248320, |
| 5120 |
| ], |
| "num_rows": 248320, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 516505600, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999805, |
| "clip_improvement": 0.250989 |
| }, |
| "model.embed_tokens": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 248320, |
| 5120 |
| ], |
| "num_rows": 248320, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 516505600, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999813, |
| "clip_improvement": 0.250535 |
| }, |
| "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": 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": 0.999844, |
| "clip_improvement": 0.250091 |
| }, |
| "model.layers.0.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": 0.999849, |
| "clip_improvement": 0.250607 |
| }, |
| "model.layers.0.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": 0.999845, |
| "clip_improvement": 0.249668 |
| }, |
| "model.layers.0.mlp.down_proj": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "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": 0.999987, |
| "clip_improvement": 0.353846 |
| }, |
| "model.layers.0.mlp.gate_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": 0.999868, |
| "clip_improvement": 0.353318 |
| }, |
| "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": 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": 0.999831, |
| "clip_improvement": 0.250704 |
| }, |
| "model.layers.1.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": 0.99987, |
| "clip_improvement": 0.250891 |
| }, |
| "model.layers.1.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": 0.999837, |
| "clip_improvement": 0.250937 |
| }, |
| "model.layers.1.mlp.down_proj": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "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": 0.999721, |
| "clip_improvement": 0.354811 |
| }, |
| "model.layers.1.mlp.gate_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": 0.999894, |
| "clip_improvement": 0.353783 |
| }, |
| "model.layers.1.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": 0.999967, |
| "clip_improvement": 0.353796 |
| }, |
| "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": 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": 0.999858, |
| "clip_improvement": 0.250787 |
| }, |
| "model.layers.10.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": 0.999849, |
| "clip_improvement": 0.250581 |
| }, |
| "model.layers.10.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": 0.999821, |
| "clip_improvement": 0.250037 |
| }, |
| "model.layers.10.mlp.down_proj": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "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": 0.999879, |
| "clip_improvement": 0.353552 |
| }, |
| "model.layers.10.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": 0.999825, |
| "clip_improvement": 0.250361 |
| }, |
| "model.layers.10.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": 0.999994, |
| "clip_improvement": 0.353712 |
| }, |
| "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": 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": 0.250154 |
| }, |
| "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": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 2129920, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999927, |
| "clip_improvement": 0.253593 |
| }, |
| "model.layers.11.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": 0.999849, |
| "clip_improvement": 0.250674 |
| }, |
| "model.layers.11.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": 0.999827, |
| "clip_improvement": 0.251097 |
| }, |
| "model.layers.11.self_attn.v_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 2129920, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999829, |
| "clip_improvement": 0.251107 |
| }, |
| "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": 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": 0.99988, |
| "clip_improvement": 0.250457 |
| }, |
| "model.layers.12.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": 0.999813, |
| "clip_improvement": 0.250526 |
| }, |
| "model.layers.12.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": 0.999825, |
| "clip_improvement": 0.250706 |
| }, |
| "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": 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": 0.250209 |
| }, |
| "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": 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": 0.999875, |
| "clip_improvement": 0.250257 |
| }, |
| "model.layers.13.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": 0.250167 |
| }, |
| "model.layers.13.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": 0.999849, |
| "clip_improvement": 0.250784 |
| }, |
| "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": 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": 0.249919 |
| }, |
| "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": 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": 0.999871, |
| "clip_improvement": 0.250805 |
| }, |
| "model.layers.14.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": 0.999805, |
| "clip_improvement": 0.250627 |
| }, |
| "model.layers.14.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": 0.999845, |
| "clip_improvement": 0.250846 |
| }, |
| "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": 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": 0.249788 |
| }, |
| "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": 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": 0.249957 |
| }, |
| "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": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 2129920, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999902, |
| "clip_improvement": 0.252909 |
| }, |
| "model.layers.15.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": 0.999841, |
| "clip_improvement": 0.250455 |
| }, |
| "model.layers.15.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": 0.99978, |
| "clip_improvement": 0.25103 |
| }, |
| "model.layers.15.self_attn.v_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 2129920, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999854, |
| "clip_improvement": 0.251013 |
| }, |
| "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": 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": 0.999846, |
| "clip_improvement": 0.250546 |
| }, |
| "model.layers.16.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": 0.999858, |
| "clip_improvement": 0.250616 |
| }, |
| "model.layers.16.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": 0.999906, |
| "clip_improvement": 0.25027 |
| }, |
| "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": 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": 0.999828, |
| "clip_improvement": 0.250176 |
| }, |
| "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": 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": 0.999834, |
| "clip_improvement": 0.250565 |
| }, |
| "model.layers.17.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": 0.999866, |
| "clip_improvement": 0.250324 |
| }, |
| "model.layers.17.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": 0.999841, |
| "clip_improvement": 0.250641 |
| }, |
| "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": 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": 0.999829, |
| "clip_improvement": 0.250166 |
| }, |
| "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": 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": 0.99979, |
| "clip_improvement": 0.250826 |
| }, |
| "model.layers.18.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": 0.999862, |
| "clip_improvement": 0.250157 |
| }, |
| "model.layers.18.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": 0.999849, |
| "clip_improvement": 0.251954 |
| }, |
| "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": 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": 0.999809, |
| "clip_improvement": 0.250923 |
| }, |
| "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": 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": 0.250005 |
| }, |
| "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": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 2129920, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999902, |
| "clip_improvement": 0.25286 |
| }, |
| "model.layers.19.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": 0.999845, |
| "clip_improvement": 0.250766 |
| }, |
| "model.layers.19.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": 0.999839, |
| "clip_improvement": 0.250839 |
| }, |
| "model.layers.19.self_attn.v_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 2129920, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999536, |
| "clip_improvement": 0.253838 |
| }, |
| "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": 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": 0.999834, |
| "clip_improvement": 0.250582 |
| }, |
| "model.layers.2.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": 0.999805, |
| "clip_improvement": 0.250406 |
| }, |
| "model.layers.2.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": 0.251606 |
| }, |
| "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": 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": 0.999961, |
| "clip_improvement": 0.353647 |
| }, |
| "model.layers.2.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": 0.999974, |
| "clip_improvement": 0.353523 |
| }, |
| "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": 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": 0.999834, |
| "clip_improvement": 0.250579 |
| }, |
| "model.layers.20.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": 0.999854, |
| "clip_improvement": 0.250433 |
| }, |
| "model.layers.20.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": 0.999849, |
| "clip_improvement": 0.251395 |
| }, |
| "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": 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": 0.999816, |
| "clip_improvement": 0.25006 |
| }, |
| "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": 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": 0.999841, |
| "clip_improvement": 0.250783 |
| }, |
| "model.layers.21.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": 0.999898, |
| "clip_improvement": 0.250342 |
| }, |
| "model.layers.21.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": 0.999817, |
| "clip_improvement": 0.250441 |
| }, |
| "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": 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": 0.249956 |
| }, |
| "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": 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": 0.999836, |
| "clip_improvement": 0.250619 |
| }, |
| "model.layers.22.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": 0.999837, |
| "clip_improvement": 0.250034 |
| }, |
| "model.layers.22.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": 0.999837, |
| "clip_improvement": 0.250538 |
| }, |
| "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": 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": 0.250545 |
| }, |
| "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": 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": 0.250251 |
| }, |
| "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": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 2129920, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999829, |
| "clip_improvement": 0.253369 |
| }, |
| "model.layers.23.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": 0.999829, |
| "clip_improvement": 0.250832 |
| }, |
| "model.layers.23.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": 0.999829, |
| "clip_improvement": 0.25062 |
| }, |
| "model.layers.23.self_attn.v_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 2129920, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999854, |
| "clip_improvement": 0.251332 |
| }, |
| "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": 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": 0.999895, |
| "clip_improvement": 0.250916 |
| }, |
| "model.layers.24.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": 0.999849, |
| "clip_improvement": 0.250412 |
| }, |
| "model.layers.24.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": 0.999862, |
| "clip_improvement": 0.250533 |
| }, |
| "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": 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": 0.250359 |
| }, |
| "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": 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": 0.999885, |
| "clip_improvement": 0.250974 |
| }, |
| "model.layers.25.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": 0.999809, |
| "clip_improvement": 0.250391 |
| }, |
| "model.layers.25.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": 0.999878, |
| "clip_improvement": 0.250444 |
| }, |
| "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": 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": 0.250066 |
| }, |
| "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": 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": 0.999866, |
| "clip_improvement": 0.251135 |
| }, |
| "model.layers.26.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": 0.999866, |
| "clip_improvement": 0.250561 |
| }, |
| "model.layers.26.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": 0.999858, |
| "clip_improvement": 0.251026 |
| }, |
| "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": 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": 0.250471 |
| }, |
| "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": 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": 0.25029 |
| }, |
| "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": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 2129920, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999878, |
| "clip_improvement": 0.254927 |
| }, |
| "model.layers.27.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": 0.999878, |
| "clip_improvement": 0.250476 |
| }, |
| "model.layers.27.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": 0.999829, |
| "clip_improvement": 0.250888 |
| }, |
| "model.layers.27.self_attn.v_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 2129920, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999683, |
| "clip_improvement": 0.25221 |
| }, |
| "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": 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": 0.999851, |
| "clip_improvement": 0.251185 |
| }, |
| "model.layers.28.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": 0.999821, |
| "clip_improvement": 0.250604 |
| }, |
| "model.layers.28.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": 0.999841, |
| "clip_improvement": 0.250859 |
| }, |
| "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": 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": 0.250338 |
| }, |
| "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": 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": 0.999866, |
| "clip_improvement": 0.250794 |
| }, |
| "model.layers.29.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": 0.999862, |
| "clip_improvement": 0.250709 |
| }, |
| "model.layers.29.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": 0.999886, |
| "clip_improvement": 0.250251 |
| }, |
| "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": 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": 0.999855, |
| "clip_improvement": 0.250481 |
| }, |
| "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": 2, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "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": 0.999974, |
| "clip_improvement": 0.353523 |
| }, |
| "model.layers.3.mlp.gate_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": 0.99999, |
| "clip_improvement": 0.353338 |
| }, |
| "model.layers.3.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": 0.999989, |
| "clip_improvement": 0.3536 |
| }, |
| "model.layers.3.self_attn.k_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 2129920, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999854, |
| "clip_improvement": 0.252481 |
| }, |
| "model.layers.3.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": 0.999845, |
| "clip_improvement": 0.249997 |
| }, |
| "model.layers.3.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": 0.999858, |
| "clip_improvement": 0.250812 |
| }, |
| "model.layers.3.self_attn.v_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 2129920, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999854, |
| "clip_improvement": 0.250395 |
| }, |
| "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": 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": 0.999853, |
| "clip_improvement": 0.251121 |
| }, |
| "model.layers.30.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": 0.999874, |
| "clip_improvement": 0.250881 |
| }, |
| "model.layers.30.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": 0.999817, |
| "clip_improvement": 0.250916 |
| }, |
| "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": 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": 0.999829, |
| "clip_improvement": 0.250478 |
| }, |
| "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": 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": 0.250606 |
| }, |
| "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": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 2129920, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999878, |
| "clip_improvement": 0.252961 |
| }, |
| "model.layers.31.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": 0.999849, |
| "clip_improvement": 0.250563 |
| }, |
| "model.layers.31.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": 0.999872, |
| "clip_improvement": 0.250712 |
| }, |
| "model.layers.31.self_attn.v_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 2129920, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999854, |
| "clip_improvement": 0.251728 |
| }, |
| "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": 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": 0.99989, |
| "clip_improvement": 0.250942 |
| }, |
| "model.layers.32.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": 0.999878, |
| "clip_improvement": 0.250956 |
| }, |
| "model.layers.32.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": 0.999845, |
| "clip_improvement": 0.250989 |
| }, |
| "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": 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": 0.250842 |
| }, |
| "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": 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": 0.999888, |
| "clip_improvement": 0.251302 |
| }, |
| "model.layers.33.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": 0.999866, |
| "clip_improvement": 0.251082 |
| }, |
| "model.layers.33.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": 0.999825, |
| "clip_improvement": 0.250962 |
| }, |
| "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": 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": 0.251086 |
| }, |
| "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": 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": 0.999878, |
| "clip_improvement": 0.251089 |
| }, |
| "model.layers.34.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": 0.999805, |
| "clip_improvement": 0.251384 |
| }, |
| "model.layers.34.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": 0.999854, |
| "clip_improvement": 0.250989 |
| }, |
| "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": 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": 0.999785, |
| "clip_improvement": 0.252622 |
| }, |
| "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": 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": 0.999819, |
| "clip_improvement": 0.250965 |
| }, |
| "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": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 2129920, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.99978, |
| "clip_improvement": 0.253111 |
| }, |
| "model.layers.35.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": 0.999862, |
| "clip_improvement": 0.250769 |
| }, |
| "model.layers.35.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": 0.999854, |
| "clip_improvement": 0.250772 |
| }, |
| "model.layers.35.self_attn.v_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 2129920, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.99978, |
| "clip_improvement": 0.253496 |
| }, |
| "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": 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": 0.999846, |
| "clip_improvement": 0.250961 |
| }, |
| "model.layers.36.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": 0.99987, |
| "clip_improvement": 0.250413 |
| }, |
| "model.layers.36.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": 0.999854, |
| "clip_improvement": 0.25111 |
| }, |
| "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": 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": 0.250866 |
| }, |
| "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": 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": 0.999675, |
| "clip_improvement": 0.251486 |
| }, |
| "model.layers.37.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": 0.999821, |
| "clip_improvement": 0.250887 |
| }, |
| "model.layers.37.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": 0.999813, |
| "clip_improvement": 0.251294 |
| }, |
| "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": 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": 0.250553 |
| }, |
| "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": 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": 0.999839, |
| "clip_improvement": 0.250733 |
| }, |
| "model.layers.38.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": 0.999862, |
| "clip_improvement": 0.250613 |
| }, |
| "model.layers.38.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": 0.999849, |
| "clip_improvement": 0.250635 |
| }, |
| "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": 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": 0.250539 |
| }, |
| "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": 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": 0.250093 |
| }, |
| "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": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 2129920, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999756, |
| "clip_improvement": 0.254395 |
| }, |
| "model.layers.39.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": 0.999841, |
| "clip_improvement": 0.250631 |
| }, |
| "model.layers.39.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": 0.999868, |
| "clip_improvement": 0.250693 |
| }, |
| "model.layers.39.self_attn.v_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 2129920, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999829, |
| "clip_improvement": 0.252474 |
| }, |
| "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": 2, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 14745600, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.999983, |
| "clip_improvement": 0.353219 |
| }, |
| "model.layers.4.linear_attn.in_proj_z": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 8847360, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.999923, |
| "clip_improvement": 0.352992 |
| }, |
| "model.layers.4.linear_attn.out_proj": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 8847360, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.999939, |
| "clip_improvement": 0.355731 |
| }, |
| "model.layers.4.mlp.down_proj": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "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": 0.999957, |
| "clip_improvement": 0.353171 |
| }, |
| "model.layers.4.mlp.gate_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": 0.999976, |
| "clip_improvement": 0.353224 |
| }, |
| "model.layers.4.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": 0.999994, |
| "clip_improvement": 0.353679 |
| }, |
| "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": 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": 0.999841, |
| "clip_improvement": 0.251388 |
| }, |
| "model.layers.40.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": 0.999793, |
| "clip_improvement": 0.250878 |
| }, |
| "model.layers.40.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": 0.999886, |
| "clip_improvement": 0.251604 |
| }, |
| "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": 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": 0.250328 |
| }, |
| "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": 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": 0.999878, |
| "clip_improvement": 0.250711 |
| }, |
| "model.layers.41.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": 0.999878, |
| "clip_improvement": 0.250691 |
| }, |
| "model.layers.41.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": 0.999845, |
| "clip_improvement": 0.250379 |
| }, |
| "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": 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": 0.999806, |
| "clip_improvement": 0.250359 |
| }, |
| "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": 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": 0.999792, |
| "clip_improvement": 0.251079 |
| }, |
| "model.layers.42.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": 0.999882, |
| "clip_improvement": 0.250732 |
| }, |
| "model.layers.42.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": 0.999805, |
| "clip_improvement": 0.251127 |
| }, |
| "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": 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": 0.999816, |
| "clip_improvement": 0.250717 |
| }, |
| "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": 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": 0.999815, |
| "clip_improvement": 0.250305 |
| }, |
| "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": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 2129920, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999805, |
| "clip_improvement": 0.254231 |
| }, |
| "model.layers.43.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": 0.999882, |
| "clip_improvement": 0.251551 |
| }, |
| "model.layers.43.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": 0.999876, |
| "clip_improvement": 0.251014 |
| }, |
| "model.layers.43.self_attn.v_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 2129920, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999731, |
| "clip_improvement": 0.255129 |
| }, |
| "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": 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": 0.999839, |
| "clip_improvement": 0.251042 |
| }, |
| "model.layers.44.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": 0.999825, |
| "clip_improvement": 0.25104 |
| }, |
| "model.layers.44.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": 0.251 |
| }, |
| "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": 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": 0.999828, |
| "clip_improvement": 0.250627 |
| }, |
| "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": 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": 0.999853, |
| "clip_improvement": 0.250857 |
| }, |
| "model.layers.45.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": 0.999821, |
| "clip_improvement": 0.250449 |
| }, |
| "model.layers.45.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": 0.99987, |
| "clip_improvement": 0.25036 |
| }, |
| "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": 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": 0.999819, |
| "clip_improvement": 0.250478 |
| }, |
| "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": 2, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 14745600, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.999744, |
| "clip_improvement": 0.354854 |
| }, |
| "model.layers.46.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": 0.999849, |
| "clip_improvement": 0.25129 |
| }, |
| "model.layers.46.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": 0.999837, |
| "clip_improvement": 0.250893 |
| }, |
| "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": 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": 0.250671 |
| }, |
| "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": 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": 0.250689 |
| }, |
| "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": 2, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 1474560, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.999048, |
| "clip_improvement": 0.353625 |
| }, |
| "model.layers.47.self_attn.o_proj": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 8847360, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.999923, |
| "clip_improvement": 0.354173 |
| }, |
| "model.layers.47.self_attn.q_proj": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 12288, |
| 5120 |
| ], |
| "num_rows": 12288, |
| "num_groups": 40, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 17694720, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.999906, |
| "clip_improvement": 0.354033 |
| }, |
| "model.layers.47.self_attn.v_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 2129920, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999829, |
| "clip_improvement": 0.251878 |
| }, |
| "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": 2, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 14745600, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.999922, |
| "clip_improvement": 0.355062 |
| }, |
| "model.layers.48.linear_attn.in_proj_z": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 8847360, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.999882, |
| "clip_improvement": 0.353672 |
| }, |
| "model.layers.48.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": 0.999845, |
| "clip_improvement": 0.251565 |
| }, |
| "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": 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": 0.999812, |
| "clip_improvement": 0.250925 |
| }, |
| "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": 2, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 14745600, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.999905, |
| "clip_improvement": 0.354578 |
| }, |
| "model.layers.49.linear_attn.in_proj_z": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 8847360, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.999963, |
| "clip_improvement": 0.353659 |
| }, |
| "model.layers.49.linear_attn.out_proj": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 8847360, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.999902, |
| "clip_improvement": 0.35385 |
| }, |
| "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": 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": 0.999759, |
| "clip_improvement": 0.251713 |
| }, |
| "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": 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": 0.999841, |
| "clip_improvement": 0.25045 |
| }, |
| "model.layers.5.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": 0.999854, |
| "clip_improvement": 0.250406 |
| }, |
| "model.layers.5.linear_attn.out_proj": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 8847360, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.999935, |
| "clip_improvement": 0.353915 |
| }, |
| "model.layers.5.mlp.down_proj": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "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": 0.999958, |
| "clip_improvement": 0.353008 |
| }, |
| "model.layers.5.mlp.gate_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": 0.999968, |
| "clip_improvement": 0.353494 |
| }, |
| "model.layers.5.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": 0.999994, |
| "clip_improvement": 0.353572 |
| }, |
| "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": 2, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 14745600, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.999585, |
| "clip_improvement": 0.354729 |
| }, |
| "model.layers.50.linear_attn.in_proj_z": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 8847360, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.999927, |
| "clip_improvement": 0.354149 |
| }, |
| "model.layers.50.linear_attn.out_proj": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 8847360, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.99974, |
| "clip_improvement": 0.355047 |
| }, |
| "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": 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": 0.999812, |
| "clip_improvement": 0.251848 |
| }, |
| "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": 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": 0.25114 |
| }, |
| "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": 2, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 1474560, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.998633, |
| "clip_improvement": 0.35126 |
| }, |
| "model.layers.51.self_attn.o_proj": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 8847360, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.999833, |
| "clip_improvement": 0.355382 |
| }, |
| "model.layers.51.self_attn.q_proj": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 12288, |
| 5120 |
| ], |
| "num_rows": 12288, |
| "num_groups": 40, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 17694720, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.99999, |
| "clip_improvement": 0.354918 |
| }, |
| "model.layers.51.self_attn.v_proj": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 1474560, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.994458, |
| "clip_improvement": 0.357812 |
| }, |
| "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": 2, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 14745600, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.999807, |
| "clip_improvement": 0.352872 |
| }, |
| "model.layers.52.linear_attn.in_proj_z": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 8847360, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.999927, |
| "clip_improvement": 0.354082 |
| }, |
| "model.layers.52.linear_attn.out_proj": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 8847360, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.999675, |
| "clip_improvement": 0.356662 |
| }, |
| "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": 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": 0.999819, |
| "clip_improvement": 0.251123 |
| }, |
| "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": 2, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 14745600, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.999705, |
| "clip_improvement": 0.35335 |
| }, |
| "model.layers.53.linear_attn.in_proj_z": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 8847360, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.999902, |
| "clip_improvement": 0.353271 |
| }, |
| "model.layers.53.linear_attn.out_proj": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 8847360, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.99967, |
| "clip_improvement": 0.355639 |
| }, |
| "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": 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": 0.999829, |
| "clip_improvement": 0.250976 |
| }, |
| "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": 2, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 14745600, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.999565, |
| "clip_improvement": 0.353618 |
| }, |
| "model.layers.54.linear_attn.in_proj_z": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 8847360, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.999976, |
| "clip_improvement": 0.353476 |
| }, |
| "model.layers.54.linear_attn.out_proj": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 8847360, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.999658, |
| "clip_improvement": 0.357367 |
| }, |
| "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": 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": 0.999855, |
| "clip_improvement": 0.250931 |
| }, |
| "model.layers.54.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": 0.999917, |
| "clip_improvement": 0.354269 |
| }, |
| "model.layers.55.mlp.down_proj": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "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": 0.999947, |
| "clip_improvement": 0.3541 |
| }, |
| "model.layers.55.mlp.gate_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": 0.999937, |
| "clip_improvement": 0.353635 |
| }, |
| "model.layers.55.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": 0.999977, |
| "clip_improvement": 0.353981 |
| }, |
| "model.layers.55.self_attn.k_proj": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 1474560, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.997241, |
| "clip_improvement": 0.347822 |
| }, |
| "model.layers.55.self_attn.o_proj": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 8847360, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.999825, |
| "clip_improvement": 0.354671 |
| }, |
| "model.layers.55.self_attn.q_proj": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 12288, |
| 5120 |
| ], |
| "num_rows": 12288, |
| "num_groups": 40, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 17694720, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.999939, |
| "clip_improvement": 0.353314 |
| }, |
| "model.layers.55.self_attn.v_proj": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 1474560, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.997168, |
| "clip_improvement": 0.353878 |
| }, |
| "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": 2, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 14745600, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.999871, |
| "clip_improvement": 0.352875 |
| }, |
| "model.layers.56.linear_attn.in_proj_z": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 8847360, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.999972, |
| "clip_improvement": 0.353063 |
| }, |
| "model.layers.56.linear_attn.out_proj": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 8847360, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.999813, |
| "clip_improvement": 0.358141 |
| }, |
| "model.layers.56.mlp.down_proj": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "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": 0.999964, |
| "clip_improvement": 0.353904 |
| }, |
| "model.layers.56.mlp.gate_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": 0.99992, |
| "clip_improvement": 0.353673 |
| }, |
| "model.layers.56.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": 0.999977, |
| "clip_improvement": 0.353969 |
| }, |
| "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": 2, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 14745600, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.999739, |
| "clip_improvement": 0.352639 |
| }, |
| "model.layers.57.linear_attn.in_proj_z": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 8847360, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.999955, |
| "clip_improvement": 0.352948 |
| }, |
| "model.layers.57.linear_attn.out_proj": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 8847360, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.999605, |
| "clip_improvement": 0.355793 |
| }, |
| "model.layers.57.mlp.down_proj": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "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": 0.999984, |
| "clip_improvement": 0.353749 |
| }, |
| "model.layers.57.mlp.gate_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": 0.999925, |
| "clip_improvement": 0.353474 |
| }, |
| "model.layers.57.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": 0.999974, |
| "clip_improvement": 0.353772 |
| }, |
| "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": 2, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 14745600, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.999749, |
| "clip_improvement": 0.352913 |
| }, |
| "model.layers.58.linear_attn.in_proj_z": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 8847360, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.999963, |
| "clip_improvement": 0.353247 |
| }, |
| "model.layers.58.linear_attn.out_proj": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 8847360, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.999837, |
| "clip_improvement": 0.357009 |
| }, |
| "model.layers.58.mlp.down_proj": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "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": 0.999977, |
| "clip_improvement": 0.354107 |
| }, |
| "model.layers.58.mlp.gate_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": 0.999854, |
| "clip_improvement": 0.353698 |
| }, |
| "model.layers.58.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": 0.999968, |
| "clip_improvement": 0.353992 |
| }, |
| "model.layers.59.mlp.down_proj": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "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": 0.999958, |
| "clip_improvement": 0.355231 |
| }, |
| "model.layers.59.mlp.gate_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": 0.999895, |
| "clip_improvement": 0.35364 |
| }, |
| "model.layers.59.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": 0.999945, |
| "clip_improvement": 0.354107 |
| }, |
| "model.layers.59.self_attn.k_proj": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 1474560, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.996924, |
| "clip_improvement": 0.347415 |
| }, |
| "model.layers.59.self_attn.o_proj": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 8847360, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.999862, |
| "clip_improvement": 0.355754 |
| }, |
| "model.layers.59.self_attn.q_proj": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 12288, |
| 5120 |
| ], |
| "num_rows": 12288, |
| "num_groups": 40, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 17694720, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.999707, |
| "clip_improvement": 0.352399 |
| }, |
| "model.layers.59.self_attn.v_proj": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 1474560, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.997217, |
| "clip_improvement": 0.358638 |
| }, |
| "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": 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": 0.999844, |
| "clip_improvement": 0.250666 |
| }, |
| "model.layers.6.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": 0.999878, |
| "clip_improvement": 0.250162 |
| }, |
| "model.layers.6.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": 0.999849, |
| "clip_improvement": 0.251858 |
| }, |
| "model.layers.6.mlp.down_proj": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "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": 0.999269, |
| "clip_improvement": 0.358459 |
| }, |
| "model.layers.6.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": 0.999792, |
| "clip_improvement": 0.250542 |
| }, |
| "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": 2, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 14745600, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.99947, |
| "clip_improvement": 0.352356 |
| }, |
| "model.layers.60.linear_attn.in_proj_z": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 8847360, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.999866, |
| "clip_improvement": 0.352588 |
| }, |
| "model.layers.60.linear_attn.out_proj": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 8847360, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.999618, |
| "clip_improvement": 0.357714 |
| }, |
| "model.layers.60.mlp.down_proj": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "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": 0.354047 |
| }, |
| "model.layers.60.mlp.gate_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": 0.999892, |
| "clip_improvement": 0.353722 |
| }, |
| "model.layers.60.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": 0.999967, |
| "clip_improvement": 0.353972 |
| }, |
| "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": 2, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 14745600, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.999663, |
| "clip_improvement": 0.352677 |
| }, |
| "model.layers.61.linear_attn.in_proj_z": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 8847360, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.999858, |
| "clip_improvement": 0.352217 |
| }, |
| "model.layers.61.linear_attn.out_proj": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 8847360, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.999776, |
| "clip_improvement": 0.354437 |
| }, |
| "model.layers.61.mlp.down_proj": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "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": 0.999966, |
| "clip_improvement": 0.354029 |
| }, |
| "model.layers.61.mlp.gate_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": 0.999869, |
| "clip_improvement": 0.353555 |
| }, |
| "model.layers.61.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": 0.999957, |
| "clip_improvement": 0.354352 |
| }, |
| "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": 2, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 10240, |
| 5120 |
| ], |
| "num_rows": 10240, |
| "num_groups": 40, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 14745600, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.99897, |
| "clip_improvement": 0.3539 |
| }, |
| "model.layers.62.linear_attn.in_proj_z": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 6144, |
| 5120 |
| ], |
| "num_rows": 6144, |
| "num_groups": 40, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 8847360, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.999548, |
| "clip_improvement": 0.352459 |
| }, |
| "model.layers.62.linear_attn.out_proj": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 8847360, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.999544, |
| "clip_improvement": 0.358339 |
| }, |
| "model.layers.62.mlp.down_proj": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "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": 0.999865, |
| "clip_improvement": 0.355188 |
| }, |
| "model.layers.62.mlp.gate_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": 0.999869, |
| "clip_improvement": 0.35315 |
| }, |
| "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": 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": 0.999799, |
| "clip_improvement": 0.353308 |
| }, |
| "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": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 2129920, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999707, |
| "clip_improvement": 0.256872 |
| }, |
| "model.layers.63.self_attn.o_proj": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 6144, |
| "logical_shape": [ |
| 5120, |
| 6144 |
| ], |
| "num_rows": 5120, |
| "num_groups": 48, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 8847360, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.999581, |
| "clip_improvement": 0.358974 |
| }, |
| "model.layers.63.self_attn.q_proj": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 12288, |
| 5120 |
| ], |
| "num_rows": 12288, |
| "num_groups": 40, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 17694720, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.999632, |
| "clip_improvement": 0.352072 |
| }, |
| "model.layers.63.self_attn.v_proj": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 8, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 1474560, |
| "bits_per_weight": 2.25, |
| "clipped_fraction": 0.995996, |
| "clip_improvement": 0.358435 |
| }, |
| "model.layers.7.mlp.down_proj": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "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": 0.999981, |
| "clip_improvement": 0.353291 |
| }, |
| "model.layers.7.mlp.gate_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": 0.353568 |
| }, |
| "model.layers.7.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": 0.999997, |
| "clip_improvement": 0.353725 |
| }, |
| "model.layers.7.self_attn.k_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 2129920, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.999902, |
| "clip_improvement": 0.252616 |
| }, |
| "model.layers.7.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": 0.999862, |
| "clip_improvement": 0.250596 |
| }, |
| "model.layers.7.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": 0.999835, |
| "clip_improvement": 0.250956 |
| }, |
| "model.layers.7.self_attn.v_proj": { |
| "bits": 3, |
| "group_size": 128, |
| "in_features": 5120, |
| "logical_shape": [ |
| 1024, |
| 5120 |
| ], |
| "num_rows": 1024, |
| "num_groups": 40, |
| "words_per_group": 12, |
| "row_offset": 0, |
| "layout": "linear", |
| "symmetric": false, |
| "has_offsets": true, |
| "compute_dtype": "bfloat16", |
| "nbytes": 2129920, |
| "bits_per_weight": 3.25, |
| "clipped_fraction": 0.99978, |
| "clip_improvement": 0.251199 |
| }, |
| "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": 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": 0.999853, |
| "clip_improvement": 0.25062 |
| }, |
| "model.layers.8.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": 0.999866, |
| "clip_improvement": 0.250399 |
| }, |
| "model.layers.8.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": 0.999874, |
| "clip_improvement": 0.25138 |
| }, |
| "model.layers.8.mlp.down_proj": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "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": 0.999991, |
| "clip_improvement": 0.353481 |
| }, |
| "model.layers.8.mlp.gate_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": 0.999983, |
| "clip_improvement": 0.353445 |
| }, |
| "model.layers.8.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": 0.999997, |
| "clip_improvement": 0.353717 |
| }, |
| "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": 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": 0.999849, |
| "clip_improvement": 0.250644 |
| }, |
| "model.layers.9.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": 0.999833, |
| "clip_improvement": 0.250352 |
| }, |
| "model.layers.9.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": 0.999849, |
| "clip_improvement": 0.250981 |
| }, |
| "model.layers.9.mlp.down_proj": { |
| "bits": 2, |
| "group_size": 128, |
| "in_features": 17408, |
| "logical_shape": [ |
| 5120, |
| 17408 |
| ], |
| "num_rows": 5120, |
| "num_groups": 136, |
| "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": 0.999977, |
| "clip_improvement": 0.353306 |
| }, |
| "model.layers.9.mlp.gate_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": 0.999951, |
| "clip_improvement": 0.35341 |
| }, |
| "model.layers.9.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": 0.999999, |
| "clip_improvement": 0.353645 |
| } |
| } |
| } |
