mixlab-gpt2-strict-small-replica/weight_map.json

[
  {
    "mixlab": "embed",
    "hf": "transformer.wte.weight",
    "shape": [
      16384,
      768
    ]
  },
  {
    "mixlab": "position_embeddings",
    "hf": "transformer.wpe.weight",
    "shape": [
      1024,
      768
    ]
  },
  {
    "mixlab": "norm_scale",
    "hf": "transformer.h.0.ln_1.weight",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "norm_bias",
    "hf": "transformer.h.0.ln_1.bias",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "wq,wk,wv",
    "hf": "transformer.h.0.attn.c_attn.weight",
    "shape": [
      768,
      2304
    ]
  },
  {
    "mixlab": "wq_bias,wk_bias,wv_bias",
    "hf": "transformer.h.0.attn.c_attn.bias",
    "shape": [
      2304
    ]
  },
  {
    "mixlab": "wo",
    "hf": "transformer.h.0.attn.c_proj.weight",
    "shape": [
      768,
      768
    ]
  },
  {
    "mixlab": "wo_bias",
    "hf": "transformer.h.0.attn.c_proj.bias",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "ffn_norm_scale",
    "hf": "transformer.h.0.ln_2.weight",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "ffn_norm_bias",
    "hf": "transformer.h.0.ln_2.bias",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "ff1",
    "hf": "transformer.h.0.mlp.c_fc.weight",
    "shape": [
      768,
      3072
    ]
  },
  {
    "mixlab": "ff1_bias",
    "hf": "transformer.h.0.mlp.c_fc.bias",
    "shape": [
      3072
    ]
  },
  {
    "mixlab": "ff2",
    "hf": "transformer.h.0.mlp.c_proj.weight",
    "shape": [
      3072,
      768
    ]
  },
  {
    "mixlab": "ff2_bias",
    "hf": "transformer.h.0.mlp.c_proj.bias",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "norm_scale",
    "hf": "transformer.h.1.ln_1.weight",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "norm_bias",
    "hf": "transformer.h.1.ln_1.bias",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "wq,wk,wv",
    "hf": "transformer.h.1.attn.c_attn.weight",
    "shape": [
      768,
      2304
    ]
  },
  {
    "mixlab": "wq_bias,wk_bias,wv_bias",
    "hf": "transformer.h.1.attn.c_attn.bias",
    "shape": [
      2304
    ]
  },
  {
    "mixlab": "wo",
    "hf": "transformer.h.1.attn.c_proj.weight",
    "shape": [
      768,
      768
    ]
  },
  {
    "mixlab": "wo_bias",
    "hf": "transformer.h.1.attn.c_proj.bias",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "ffn_norm_scale",
    "hf": "transformer.h.1.ln_2.weight",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "ffn_norm_bias",
    "hf": "transformer.h.1.ln_2.bias",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "ff1",
    "hf": "transformer.h.1.mlp.c_fc.weight",
    "shape": [
      768,
      3072
    ]
  },
  {
    "mixlab": "ff1_bias",
    "hf": "transformer.h.1.mlp.c_fc.bias",
    "shape": [
      3072
    ]
  },
  {
    "mixlab": "ff2",
    "hf": "transformer.h.1.mlp.c_proj.weight",
    "shape": [
      3072,
      768
    ]
  },
  {
    "mixlab": "ff2_bias",
    "hf": "transformer.h.1.mlp.c_proj.bias",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "norm_scale",
    "hf": "transformer.h.2.ln_1.weight",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "norm_bias",
    "hf": "transformer.h.2.ln_1.bias",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "wq,wk,wv",
    "hf": "transformer.h.2.attn.c_attn.weight",
    "shape": [
      768,
      2304
    ]
  },
  {
    "mixlab": "wq_bias,wk_bias,wv_bias",
    "hf": "transformer.h.2.attn.c_attn.bias",
    "shape": [
      2304
    ]
  },
  {
    "mixlab": "wo",
    "hf": "transformer.h.2.attn.c_proj.weight",
    "shape": [
      768,
      768
    ]
  },
  {
    "mixlab": "wo_bias",
    "hf": "transformer.h.2.attn.c_proj.bias",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "ffn_norm_scale",
    "hf": "transformer.h.2.ln_2.weight",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "ffn_norm_bias",
    "hf": "transformer.h.2.ln_2.bias",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "ff1",
    "hf": "transformer.h.2.mlp.c_fc.weight",
    "shape": [
      768,
      3072
    ]
  },
  {
    "mixlab": "ff1_bias",
    "hf": "transformer.h.2.mlp.c_fc.bias",
    "shape": [
      3072
    ]
  },
  {
    "mixlab": "ff2",
    "hf": "transformer.h.2.mlp.c_proj.weight",
    "shape": [
      3072,
      768
    ]
  },
  {
    "mixlab": "ff2_bias",
    "hf": "transformer.h.2.mlp.c_proj.bias",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "norm_scale",
    "hf": "transformer.h.3.ln_1.weight",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "norm_bias",
    "hf": "transformer.h.3.ln_1.bias",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "wq,wk,wv",
    "hf": "transformer.h.3.attn.c_attn.weight",
    "shape": [
      768,
      2304
    ]
  },
  {
    "mixlab": "wq_bias,wk_bias,wv_bias",
    "hf": "transformer.h.3.attn.c_attn.bias",
    "shape": [
      2304
    ]
  },
  {
    "mixlab": "wo",
    "hf": "transformer.h.3.attn.c_proj.weight",
    "shape": [
      768,
      768
    ]
  },
  {
    "mixlab": "wo_bias",
    "hf": "transformer.h.3.attn.c_proj.bias",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "ffn_norm_scale",
    "hf": "transformer.h.3.ln_2.weight",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "ffn_norm_bias",
    "hf": "transformer.h.3.ln_2.bias",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "ff1",
    "hf": "transformer.h.3.mlp.c_fc.weight",
    "shape": [
      768,
      3072
    ]
  },
  {
    "mixlab": "ff1_bias",
    "hf": "transformer.h.3.mlp.c_fc.bias",
    "shape": [
      3072
    ]
  },
  {
    "mixlab": "ff2",
    "hf": "transformer.h.3.mlp.c_proj.weight",
    "shape": [
      3072,
      768
    ]
  },
  {
    "mixlab": "ff2_bias",
    "hf": "transformer.h.3.mlp.c_proj.bias",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "norm_scale",
    "hf": "transformer.h.4.ln_1.weight",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "norm_bias",
    "hf": "transformer.h.4.ln_1.bias",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "wq,wk,wv",
    "hf": "transformer.h.4.attn.c_attn.weight",
    "shape": [
      768,
      2304
    ]
  },
  {
    "mixlab": "wq_bias,wk_bias,wv_bias",
    "hf": "transformer.h.4.attn.c_attn.bias",
    "shape": [
      2304
    ]
  },
  {
    "mixlab": "wo",
    "hf": "transformer.h.4.attn.c_proj.weight",
    "shape": [
      768,
      768
    ]
  },
  {
    "mixlab": "wo_bias",
    "hf": "transformer.h.4.attn.c_proj.bias",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "ffn_norm_scale",
    "hf": "transformer.h.4.ln_2.weight",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "ffn_norm_bias",
    "hf": "transformer.h.4.ln_2.bias",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "ff1",
    "hf": "transformer.h.4.mlp.c_fc.weight",
    "shape": [
      768,
      3072
    ]
  },
  {
    "mixlab": "ff1_bias",
    "hf": "transformer.h.4.mlp.c_fc.bias",
    "shape": [
      3072
    ]
  },
  {
    "mixlab": "ff2",
    "hf": "transformer.h.4.mlp.c_proj.weight",
    "shape": [
      3072,
      768
    ]
  },
  {
    "mixlab": "ff2_bias",
    "hf": "transformer.h.4.mlp.c_proj.bias",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "norm_scale",
    "hf": "transformer.h.5.ln_1.weight",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "norm_bias",
    "hf": "transformer.h.5.ln_1.bias",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "wq,wk,wv",
    "hf": "transformer.h.5.attn.c_attn.weight",
    "shape": [
      768,
      2304
    ]
  },
  {
    "mixlab": "wq_bias,wk_bias,wv_bias",
    "hf": "transformer.h.5.attn.c_attn.bias",
    "shape": [
      2304
    ]
  },
  {
    "mixlab": "wo",
    "hf": "transformer.h.5.attn.c_proj.weight",
    "shape": [
      768,
      768
    ]
  },
  {
    "mixlab": "wo_bias",
    "hf": "transformer.h.5.attn.c_proj.bias",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "ffn_norm_scale",
    "hf": "transformer.h.5.ln_2.weight",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "ffn_norm_bias",
    "hf": "transformer.h.5.ln_2.bias",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "ff1",
    "hf": "transformer.h.5.mlp.c_fc.weight",
    "shape": [
      768,
      3072
    ]
  },
  {
    "mixlab": "ff1_bias",
    "hf": "transformer.h.5.mlp.c_fc.bias",
    "shape": [
      3072
    ]
  },
  {
    "mixlab": "ff2",
    "hf": "transformer.h.5.mlp.c_proj.weight",
    "shape": [
      3072,
      768
    ]
  },
  {
    "mixlab": "ff2_bias",
    "hf": "transformer.h.5.mlp.c_proj.bias",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "norm_scale",
    "hf": "transformer.h.6.ln_1.weight",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "norm_bias",
    "hf": "transformer.h.6.ln_1.bias",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "wq,wk,wv",
    "hf": "transformer.h.6.attn.c_attn.weight",
    "shape": [
      768,
      2304
    ]
  },
  {
    "mixlab": "wq_bias,wk_bias,wv_bias",
    "hf": "transformer.h.6.attn.c_attn.bias",
    "shape": [
      2304
    ]
  },
  {
    "mixlab": "wo",
    "hf": "transformer.h.6.attn.c_proj.weight",
    "shape": [
      768,
      768
    ]
  },
  {
    "mixlab": "wo_bias",
    "hf": "transformer.h.6.attn.c_proj.bias",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "ffn_norm_scale",
    "hf": "transformer.h.6.ln_2.weight",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "ffn_norm_bias",
    "hf": "transformer.h.6.ln_2.bias",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "ff1",
    "hf": "transformer.h.6.mlp.c_fc.weight",
    "shape": [
      768,
      3072
    ]
  },
  {
    "mixlab": "ff1_bias",
    "hf": "transformer.h.6.mlp.c_fc.bias",
    "shape": [
      3072
    ]
  },
  {
    "mixlab": "ff2",
    "hf": "transformer.h.6.mlp.c_proj.weight",
    "shape": [
      3072,
      768
    ]
  },
  {
    "mixlab": "ff2_bias",
    "hf": "transformer.h.6.mlp.c_proj.bias",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "norm_scale",
    "hf": "transformer.h.7.ln_1.weight",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "norm_bias",
    "hf": "transformer.h.7.ln_1.bias",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "wq,wk,wv",
    "hf": "transformer.h.7.attn.c_attn.weight",
    "shape": [
      768,
      2304
    ]
  },
  {
    "mixlab": "wq_bias,wk_bias,wv_bias",
    "hf": "transformer.h.7.attn.c_attn.bias",
    "shape": [
      2304
    ]
  },
  {
    "mixlab": "wo",
    "hf": "transformer.h.7.attn.c_proj.weight",
    "shape": [
      768,
      768
    ]
  },
  {
    "mixlab": "wo_bias",
    "hf": "transformer.h.7.attn.c_proj.bias",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "ffn_norm_scale",
    "hf": "transformer.h.7.ln_2.weight",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "ffn_norm_bias",
    "hf": "transformer.h.7.ln_2.bias",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "ff1",
    "hf": "transformer.h.7.mlp.c_fc.weight",
    "shape": [
      768,
      3072
    ]
  },
  {
    "mixlab": "ff1_bias",
    "hf": "transformer.h.7.mlp.c_fc.bias",
    "shape": [
      3072
    ]
  },
  {
    "mixlab": "ff2",
    "hf": "transformer.h.7.mlp.c_proj.weight",
    "shape": [
      3072,
      768
    ]
  },
  {
    "mixlab": "ff2_bias",
    "hf": "transformer.h.7.mlp.c_proj.bias",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "norm_scale",
    "hf": "transformer.h.8.ln_1.weight",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "norm_bias",
    "hf": "transformer.h.8.ln_1.bias",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "wq,wk,wv",
    "hf": "transformer.h.8.attn.c_attn.weight",
    "shape": [
      768,
      2304
    ]
  },
  {
    "mixlab": "wq_bias,wk_bias,wv_bias",
    "hf": "transformer.h.8.attn.c_attn.bias",
    "shape": [
      2304
    ]
  },
  {
    "mixlab": "wo",
    "hf": "transformer.h.8.attn.c_proj.weight",
    "shape": [
      768,
      768
    ]
  },
  {
    "mixlab": "wo_bias",
    "hf": "transformer.h.8.attn.c_proj.bias",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "ffn_norm_scale",
    "hf": "transformer.h.8.ln_2.weight",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "ffn_norm_bias",
    "hf": "transformer.h.8.ln_2.bias",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "ff1",
    "hf": "transformer.h.8.mlp.c_fc.weight",
    "shape": [
      768,
      3072
    ]
  },
  {
    "mixlab": "ff1_bias",
    "hf": "transformer.h.8.mlp.c_fc.bias",
    "shape": [
      3072
    ]
  },
  {
    "mixlab": "ff2",
    "hf": "transformer.h.8.mlp.c_proj.weight",
    "shape": [
      3072,
      768
    ]
  },
  {
    "mixlab": "ff2_bias",
    "hf": "transformer.h.8.mlp.c_proj.bias",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "norm_scale",
    "hf": "transformer.h.9.ln_1.weight",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "norm_bias",
    "hf": "transformer.h.9.ln_1.bias",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "wq,wk,wv",
    "hf": "transformer.h.9.attn.c_attn.weight",
    "shape": [
      768,
      2304
    ]
  },
  {
    "mixlab": "wq_bias,wk_bias,wv_bias",
    "hf": "transformer.h.9.attn.c_attn.bias",
    "shape": [
      2304
    ]
  },
  {
    "mixlab": "wo",
    "hf": "transformer.h.9.attn.c_proj.weight",
    "shape": [
      768,
      768
    ]
  },
  {
    "mixlab": "wo_bias",
    "hf": "transformer.h.9.attn.c_proj.bias",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "ffn_norm_scale",
    "hf": "transformer.h.9.ln_2.weight",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "ffn_norm_bias",
    "hf": "transformer.h.9.ln_2.bias",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "ff1",
    "hf": "transformer.h.9.mlp.c_fc.weight",
    "shape": [
      768,
      3072
    ]
  },
  {
    "mixlab": "ff1_bias",
    "hf": "transformer.h.9.mlp.c_fc.bias",
    "shape": [
      3072
    ]
  },
  {
    "mixlab": "ff2",
    "hf": "transformer.h.9.mlp.c_proj.weight",
    "shape": [
      3072,
      768
    ]
  },
  {
    "mixlab": "ff2_bias",
    "hf": "transformer.h.9.mlp.c_proj.bias",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "norm_scale",
    "hf": "transformer.h.10.ln_1.weight",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "norm_bias",
    "hf": "transformer.h.10.ln_1.bias",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "wq,wk,wv",
    "hf": "transformer.h.10.attn.c_attn.weight",
    "shape": [
      768,
      2304
    ]
  },
  {
    "mixlab": "wq_bias,wk_bias,wv_bias",
    "hf": "transformer.h.10.attn.c_attn.bias",
    "shape": [
      2304
    ]
  },
  {
    "mixlab": "wo",
    "hf": "transformer.h.10.attn.c_proj.weight",
    "shape": [
      768,
      768
    ]
  },
  {
    "mixlab": "wo_bias",
    "hf": "transformer.h.10.attn.c_proj.bias",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "ffn_norm_scale",
    "hf": "transformer.h.10.ln_2.weight",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "ffn_norm_bias",
    "hf": "transformer.h.10.ln_2.bias",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "ff1",
    "hf": "transformer.h.10.mlp.c_fc.weight",
    "shape": [
      768,
      3072
    ]
  },
  {
    "mixlab": "ff1_bias",
    "hf": "transformer.h.10.mlp.c_fc.bias",
    "shape": [
      3072
    ]
  },
  {
    "mixlab": "ff2",
    "hf": "transformer.h.10.mlp.c_proj.weight",
    "shape": [
      3072,
      768
    ]
  },
  {
    "mixlab": "ff2_bias",
    "hf": "transformer.h.10.mlp.c_proj.bias",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "norm_scale",
    "hf": "transformer.h.11.ln_1.weight",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "norm_bias",
    "hf": "transformer.h.11.ln_1.bias",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "wq,wk,wv",
    "hf": "transformer.h.11.attn.c_attn.weight",
    "shape": [
      768,
      2304
    ]
  },
  {
    "mixlab": "wq_bias,wk_bias,wv_bias",
    "hf": "transformer.h.11.attn.c_attn.bias",
    "shape": [
      2304
    ]
  },
  {
    "mixlab": "wo",
    "hf": "transformer.h.11.attn.c_proj.weight",
    "shape": [
      768,
      768
    ]
  },
  {
    "mixlab": "wo_bias",
    "hf": "transformer.h.11.attn.c_proj.bias",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "ffn_norm_scale",
    "hf": "transformer.h.11.ln_2.weight",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "ffn_norm_bias",
    "hf": "transformer.h.11.ln_2.bias",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "ff1",
    "hf": "transformer.h.11.mlp.c_fc.weight",
    "shape": [
      768,
      3072
    ]
  },
  {
    "mixlab": "ff1_bias",
    "hf": "transformer.h.11.mlp.c_fc.bias",
    "shape": [
      3072
    ]
  },
  {
    "mixlab": "ff2",
    "hf": "transformer.h.11.mlp.c_proj.weight",
    "shape": [
      3072,
      768
    ]
  },
  {
    "mixlab": "ff2_bias",
    "hf": "transformer.h.11.mlp.c_proj.bias",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "final_norm_scale",
    "hf": "transformer.ln_f.weight",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "final_norm_bias",
    "hf": "transformer.ln_f.bias",
    "shape": [
      768
    ]
  },
  {
    "mixlab": "embed",
    "hf": "lm_head.weight",
    "shape": [
      16384,
      768
    ]
  }
]