| { |
| "source_checkpoint": "base-anima-unify-k1-2-4-8-1024-gradacc32-step400", |
| "source_files": [ |
| { |
| "path": "diffusion_pytorch_model-00001-of-00002.safetensors", |
| "sha256": "67876d738ed881500813f64862c465a1c9f8afc82c31d17292fb3ce28eb9e3b1", |
| "size": 4983486068 |
| }, |
| { |
| "path": "diffusion_pytorch_model-00002-of-00002.safetensors", |
| "sha256": "279eb27150ecb6f82dbbf668e0fd261e3c469140acbe241d23daf0948c014552", |
| "size": 2909318512 |
| } |
| ], |
| "source_config_sha256": "3f545fd7a81e71e6cfffc558f01fb18aa363073bc6e2e2b579d4ab20327c461c", |
| "source_implementation_sha256": "ab8f926bd1034ca750baff03596ca128f15494fe2f2ba542aa4150dee891a560", |
| "dtype": "bfloat16", |
| "tensor_count": 571, |
| "tensors": { |
| "patch_embed.proj.weight": { |
| "shape": [ |
| 2048, |
| 68 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "time_embed.k_embed.norm.weight": { |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "time_embed.k_embed.t_embedder.linear_1.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "time_embed.k_embed.t_embedder.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "time_embed.k_scale": { |
| "shape": [], |
| "dtype": "torch.bfloat16" |
| }, |
| "time_embed.norm.weight": { |
| "shape": [ |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "time_embed.t_embedder.linear_1.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "time_embed.t_embedder.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.0.attn1.norm_k.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.0.attn1.norm_q.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.0.attn1.to_k.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.0.attn1.to_out.0.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.0.attn1.to_q.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.0.attn1.to_v.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.0.attn2.norm_k.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.0.attn2.norm_q.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.0.attn2.to_k.weight": { |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.0.attn2.to_out.0.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.0.attn2.to_q.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.0.attn2.to_v.weight": { |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.0.ff.net.0.proj.weight": { |
| "shape": [ |
| 8192, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.0.ff.net.2.weight": { |
| "shape": [ |
| 2048, |
| 8192 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.0.norm1.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.0.norm1.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.0.norm2.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.0.norm2.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.0.norm3.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.0.norm3.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.1.attn1.norm_k.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.1.attn1.norm_q.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.1.attn1.to_k.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.1.attn1.to_out.0.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.1.attn1.to_q.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.1.attn1.to_v.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.1.attn2.norm_k.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.1.attn2.norm_q.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.1.attn2.to_k.weight": { |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.1.attn2.to_out.0.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.1.attn2.to_q.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.1.attn2.to_v.weight": { |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.1.ff.net.0.proj.weight": { |
| "shape": [ |
| 8192, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.1.ff.net.2.weight": { |
| "shape": [ |
| 2048, |
| 8192 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.1.norm1.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.1.norm1.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.1.norm2.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.1.norm2.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.1.norm3.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.1.norm3.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.10.attn1.norm_k.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.10.attn1.norm_q.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.10.attn1.to_k.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.10.attn1.to_out.0.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.10.attn1.to_q.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.10.attn1.to_v.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.10.attn2.norm_k.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.10.attn2.norm_q.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.10.attn2.to_k.weight": { |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.10.attn2.to_out.0.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.10.attn2.to_q.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.10.attn2.to_v.weight": { |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.10.ff.net.0.proj.weight": { |
| "shape": [ |
| 8192, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.10.ff.net.2.weight": { |
| "shape": [ |
| 2048, |
| 8192 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.10.norm1.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.10.norm1.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.10.norm2.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.10.norm2.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.10.norm3.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.10.norm3.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.11.attn1.norm_k.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.11.attn1.norm_q.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.11.attn1.to_k.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.11.attn1.to_out.0.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.11.attn1.to_q.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.11.attn1.to_v.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.11.attn2.norm_k.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.11.attn2.norm_q.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.11.attn2.to_k.weight": { |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.11.attn2.to_out.0.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.11.attn2.to_q.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.11.attn2.to_v.weight": { |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.11.ff.net.0.proj.weight": { |
| "shape": [ |
| 8192, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.11.ff.net.2.weight": { |
| "shape": [ |
| 2048, |
| 8192 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.11.norm1.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.11.norm1.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.11.norm2.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.11.norm2.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.11.norm3.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.11.norm3.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.12.attn1.norm_k.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.12.attn1.norm_q.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.12.attn1.to_k.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.12.attn1.to_out.0.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.12.attn1.to_q.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.12.attn1.to_v.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.12.attn2.norm_k.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.12.attn2.norm_q.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.12.attn2.to_k.weight": { |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.12.attn2.to_out.0.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.12.attn2.to_q.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.12.attn2.to_v.weight": { |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.12.ff.net.0.proj.weight": { |
| "shape": [ |
| 8192, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.12.ff.net.2.weight": { |
| "shape": [ |
| 2048, |
| 8192 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.12.norm1.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.12.norm1.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.12.norm2.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.12.norm2.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.12.norm3.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.12.norm3.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.13.attn1.norm_k.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.13.attn1.norm_q.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.13.attn1.to_k.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.13.attn1.to_out.0.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.13.attn1.to_q.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.13.attn1.to_v.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.13.attn2.norm_k.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.13.attn2.norm_q.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.13.attn2.to_k.weight": { |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.13.attn2.to_out.0.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.13.attn2.to_q.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.13.attn2.to_v.weight": { |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.13.ff.net.0.proj.weight": { |
| "shape": [ |
| 8192, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.13.ff.net.2.weight": { |
| "shape": [ |
| 2048, |
| 8192 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.13.norm1.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.13.norm1.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.13.norm2.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.13.norm2.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.13.norm3.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.13.norm3.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.14.attn1.norm_k.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.14.attn1.norm_q.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.14.attn1.to_k.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.14.attn1.to_out.0.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.14.attn1.to_q.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.14.attn1.to_v.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.14.attn2.norm_k.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.14.attn2.norm_q.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.14.attn2.to_k.weight": { |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.14.attn2.to_out.0.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.14.attn2.to_q.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.14.attn2.to_v.weight": { |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.14.ff.net.0.proj.weight": { |
| "shape": [ |
| 8192, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.14.ff.net.2.weight": { |
| "shape": [ |
| 2048, |
| 8192 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.14.norm1.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.14.norm1.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.14.norm2.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.14.norm2.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.14.norm3.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.14.norm3.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.15.attn1.norm_k.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.15.attn1.norm_q.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.15.attn1.to_k.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.15.attn1.to_out.0.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.15.attn1.to_q.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.15.attn1.to_v.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.15.attn2.norm_k.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.15.attn2.norm_q.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.15.attn2.to_k.weight": { |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.15.attn2.to_out.0.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.15.attn2.to_q.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.15.attn2.to_v.weight": { |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.15.ff.net.0.proj.weight": { |
| "shape": [ |
| 8192, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.15.ff.net.2.weight": { |
| "shape": [ |
| 2048, |
| 8192 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.15.norm1.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.15.norm1.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.15.norm2.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.15.norm2.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.15.norm3.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.15.norm3.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.16.attn1.norm_k.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.16.attn1.norm_q.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.16.attn1.to_k.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.16.attn1.to_out.0.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.16.attn1.to_q.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.16.attn1.to_v.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.16.attn2.norm_k.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.16.attn2.norm_q.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.16.attn2.to_k.weight": { |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.16.attn2.to_out.0.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.16.attn2.to_q.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.16.attn2.to_v.weight": { |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.16.ff.net.0.proj.weight": { |
| "shape": [ |
| 8192, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.16.ff.net.2.weight": { |
| "shape": [ |
| 2048, |
| 8192 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.16.norm1.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.16.norm1.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.16.norm2.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.16.norm2.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.16.norm3.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.16.norm3.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.17.attn1.norm_k.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.17.attn1.norm_q.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.17.attn1.to_k.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.17.attn1.to_out.0.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.17.attn1.to_q.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.17.attn1.to_v.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.17.attn2.norm_k.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.17.attn2.norm_q.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.17.attn2.to_k.weight": { |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.17.attn2.to_out.0.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.17.attn2.to_q.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.17.attn2.to_v.weight": { |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.17.norm1.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.17.norm1.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.17.norm2.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.17.norm2.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.17.norm3.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.17.norm3.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.2.attn1.norm_k.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.2.attn1.norm_q.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.2.attn1.to_k.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.2.attn1.to_out.0.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.2.attn1.to_q.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.2.attn1.to_v.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.2.attn2.norm_k.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.2.attn2.norm_q.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.2.attn2.to_k.weight": { |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.2.attn2.to_out.0.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.2.attn2.to_q.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.2.attn2.to_v.weight": { |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.2.ff.net.0.proj.weight": { |
| "shape": [ |
| 8192, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.2.ff.net.2.weight": { |
| "shape": [ |
| 2048, |
| 8192 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.2.norm1.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.2.norm1.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.2.norm2.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.2.norm2.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.2.norm3.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.2.norm3.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.3.attn1.norm_k.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.3.attn1.norm_q.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.3.attn1.to_k.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.3.attn1.to_out.0.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.3.attn1.to_q.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.3.attn1.to_v.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.3.attn2.norm_k.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.3.attn2.norm_q.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.3.attn2.to_k.weight": { |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.3.attn2.to_out.0.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.3.attn2.to_q.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.3.attn2.to_v.weight": { |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.3.ff.net.0.proj.weight": { |
| "shape": [ |
| 8192, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.3.ff.net.2.weight": { |
| "shape": [ |
| 2048, |
| 8192 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.3.norm1.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.3.norm1.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.3.norm2.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.3.norm2.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.3.norm3.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.3.norm3.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.4.attn1.norm_k.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.4.attn1.norm_q.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.4.attn1.to_k.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.4.attn1.to_out.0.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.4.attn1.to_q.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.4.attn1.to_v.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.4.attn2.norm_k.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.4.attn2.norm_q.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.4.attn2.to_k.weight": { |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.4.attn2.to_out.0.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.4.attn2.to_q.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.4.attn2.to_v.weight": { |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.4.ff.net.0.proj.weight": { |
| "shape": [ |
| 8192, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.4.ff.net.2.weight": { |
| "shape": [ |
| 2048, |
| 8192 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.4.norm1.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.4.norm1.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.4.norm2.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.4.norm2.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.4.norm3.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.4.norm3.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.5.attn1.norm_k.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.5.attn1.norm_q.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.5.attn1.to_k.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.5.attn1.to_out.0.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.5.attn1.to_q.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.5.attn1.to_v.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.5.attn2.norm_k.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.5.attn2.norm_q.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.5.attn2.to_k.weight": { |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.5.attn2.to_out.0.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.5.attn2.to_q.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.5.attn2.to_v.weight": { |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.5.ff.net.0.proj.weight": { |
| "shape": [ |
| 8192, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.5.ff.net.2.weight": { |
| "shape": [ |
| 2048, |
| 8192 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.5.norm1.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.5.norm1.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.5.norm2.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.5.norm2.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.5.norm3.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.5.norm3.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.6.attn1.norm_k.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.6.attn1.norm_q.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.6.attn1.to_k.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.6.attn1.to_out.0.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.6.attn1.to_q.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.6.attn1.to_v.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.6.attn2.norm_k.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.6.attn2.norm_q.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.6.attn2.to_k.weight": { |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.6.attn2.to_out.0.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.6.attn2.to_q.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.6.attn2.to_v.weight": { |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.6.ff.net.0.proj.weight": { |
| "shape": [ |
| 8192, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.6.ff.net.2.weight": { |
| "shape": [ |
| 2048, |
| 8192 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.6.norm1.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.6.norm1.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.6.norm2.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.6.norm2.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.6.norm3.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.6.norm3.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.7.attn1.norm_k.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.7.attn1.norm_q.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.7.attn1.to_k.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.7.attn1.to_out.0.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.7.attn1.to_q.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.7.attn1.to_v.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.7.attn2.norm_k.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.7.attn2.norm_q.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.7.attn2.to_k.weight": { |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.7.attn2.to_out.0.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.7.attn2.to_q.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.7.attn2.to_v.weight": { |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.7.ff.net.0.proj.weight": { |
| "shape": [ |
| 8192, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.7.ff.net.2.weight": { |
| "shape": [ |
| 2048, |
| 8192 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.7.norm1.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.7.norm1.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.7.norm2.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.7.norm2.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.7.norm3.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.7.norm3.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.8.attn1.norm_k.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.8.attn1.norm_q.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.8.attn1.to_k.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.8.attn1.to_out.0.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.8.attn1.to_q.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.8.attn1.to_v.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.8.attn2.norm_k.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.8.attn2.norm_q.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.8.attn2.to_k.weight": { |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.8.attn2.to_out.0.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.8.attn2.to_q.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.8.attn2.to_v.weight": { |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.8.ff.net.0.proj.weight": { |
| "shape": [ |
| 8192, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.8.ff.net.2.weight": { |
| "shape": [ |
| 2048, |
| 8192 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.8.norm1.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.8.norm1.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.8.norm2.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.8.norm2.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.8.norm3.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.8.norm3.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.9.attn1.norm_k.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.9.attn1.norm_q.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.9.attn1.to_k.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.9.attn1.to_out.0.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.9.attn1.to_q.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.9.attn1.to_v.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.9.attn2.norm_k.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.9.attn2.norm_q.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.9.attn2.to_k.weight": { |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.9.attn2.to_out.0.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.9.attn2.to_q.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.9.attn2.to_v.weight": { |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.9.ff.net.0.proj.weight": { |
| "shape": [ |
| 8192, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.9.ff.net.2.weight": { |
| "shape": [ |
| 2048, |
| 8192 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.9.norm1.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.9.norm1.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.9.norm2.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.9.norm2.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.9.norm3.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.9.norm3.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "norm_out.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "norm_out.linear_2.weight": { |
| "shape": [ |
| 4096, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "proj_out.weight": { |
| "shape": [ |
| 64, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.17.ff.net.0.proj.weight": { |
| "shape": [ |
| 8192, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.17.ff.net.2.weight": { |
| "shape": [ |
| 2048, |
| 8192 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.18.attn1.norm_k.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.18.attn1.norm_q.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.18.attn1.to_k.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.18.attn1.to_out.0.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.18.attn1.to_q.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.18.attn1.to_v.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.18.attn2.norm_k.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.18.attn2.norm_q.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.18.attn2.to_k.weight": { |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.18.attn2.to_out.0.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.18.attn2.to_q.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.18.attn2.to_v.weight": { |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.18.ff.net.0.proj.weight": { |
| "shape": [ |
| 8192, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.18.ff.net.2.weight": { |
| "shape": [ |
| 2048, |
| 8192 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.18.norm1.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.18.norm1.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.18.norm2.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.18.norm2.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.18.norm3.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.18.norm3.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.19.attn1.norm_k.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.19.attn1.norm_q.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.19.attn1.to_k.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.19.attn1.to_out.0.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.19.attn1.to_q.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.19.attn1.to_v.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.19.attn2.norm_k.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.19.attn2.norm_q.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.19.attn2.to_k.weight": { |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.19.attn2.to_out.0.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.19.attn2.to_q.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.19.attn2.to_v.weight": { |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.19.ff.net.0.proj.weight": { |
| "shape": [ |
| 8192, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.19.ff.net.2.weight": { |
| "shape": [ |
| 2048, |
| 8192 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.19.norm1.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.19.norm1.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.19.norm2.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.19.norm2.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.19.norm3.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.19.norm3.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.20.attn1.norm_k.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.20.attn1.norm_q.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.20.attn1.to_k.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.20.attn1.to_out.0.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.20.attn1.to_q.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.20.attn1.to_v.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.20.attn2.norm_k.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.20.attn2.norm_q.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.20.attn2.to_k.weight": { |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.20.attn2.to_out.0.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.20.attn2.to_q.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.20.attn2.to_v.weight": { |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.20.ff.net.0.proj.weight": { |
| "shape": [ |
| 8192, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.20.ff.net.2.weight": { |
| "shape": [ |
| 2048, |
| 8192 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.20.norm1.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.20.norm1.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.20.norm2.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.20.norm2.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.20.norm3.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.20.norm3.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.21.attn1.norm_k.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.21.attn1.norm_q.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.21.attn1.to_k.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.21.attn1.to_out.0.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.21.attn1.to_q.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.21.attn1.to_v.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.21.attn2.norm_k.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.21.attn2.norm_q.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.21.attn2.to_k.weight": { |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.21.attn2.to_out.0.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.21.attn2.to_q.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.21.attn2.to_v.weight": { |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.21.ff.net.0.proj.weight": { |
| "shape": [ |
| 8192, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.21.ff.net.2.weight": { |
| "shape": [ |
| 2048, |
| 8192 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.21.norm1.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.21.norm1.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.21.norm2.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.21.norm2.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.21.norm3.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.21.norm3.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.22.attn1.norm_k.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.22.attn1.norm_q.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.22.attn1.to_k.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.22.attn1.to_out.0.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.22.attn1.to_q.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.22.attn1.to_v.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.22.attn2.norm_k.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.22.attn2.norm_q.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.22.attn2.to_k.weight": { |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.22.attn2.to_out.0.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.22.attn2.to_q.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.22.attn2.to_v.weight": { |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.22.ff.net.0.proj.weight": { |
| "shape": [ |
| 8192, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.22.ff.net.2.weight": { |
| "shape": [ |
| 2048, |
| 8192 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.22.norm1.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.22.norm1.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.22.norm2.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.22.norm2.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.22.norm3.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.22.norm3.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.23.attn1.norm_k.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.23.attn1.norm_q.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.23.attn1.to_k.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.23.attn1.to_out.0.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.23.attn1.to_q.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.23.attn1.to_v.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.23.attn2.norm_k.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.23.attn2.norm_q.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.23.attn2.to_k.weight": { |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.23.attn2.to_out.0.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.23.attn2.to_q.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.23.attn2.to_v.weight": { |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.23.ff.net.0.proj.weight": { |
| "shape": [ |
| 8192, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.23.ff.net.2.weight": { |
| "shape": [ |
| 2048, |
| 8192 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.23.norm1.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.23.norm1.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.23.norm2.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.23.norm2.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.23.norm3.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.23.norm3.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.24.attn1.norm_k.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.24.attn1.norm_q.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.24.attn1.to_k.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.24.attn1.to_out.0.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.24.attn1.to_q.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.24.attn1.to_v.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.24.attn2.norm_k.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.24.attn2.norm_q.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.24.attn2.to_k.weight": { |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.24.attn2.to_out.0.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.24.attn2.to_q.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.24.attn2.to_v.weight": { |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.24.ff.net.0.proj.weight": { |
| "shape": [ |
| 8192, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.24.ff.net.2.weight": { |
| "shape": [ |
| 2048, |
| 8192 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.24.norm1.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.24.norm1.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.24.norm2.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.24.norm2.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.24.norm3.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.24.norm3.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.25.attn1.norm_k.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.25.attn1.norm_q.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.25.attn1.to_k.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.25.attn1.to_out.0.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.25.attn1.to_q.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.25.attn1.to_v.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.25.attn2.norm_k.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.25.attn2.norm_q.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.25.attn2.to_k.weight": { |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.25.attn2.to_out.0.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.25.attn2.to_q.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.25.attn2.to_v.weight": { |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.25.ff.net.0.proj.weight": { |
| "shape": [ |
| 8192, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.25.ff.net.2.weight": { |
| "shape": [ |
| 2048, |
| 8192 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.25.norm1.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.25.norm1.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.25.norm2.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.25.norm2.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.25.norm3.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.25.norm3.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.26.attn1.norm_k.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.26.attn1.norm_q.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.26.attn1.to_k.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.26.attn1.to_out.0.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.26.attn1.to_q.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.26.attn1.to_v.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.26.attn2.norm_k.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.26.attn2.norm_q.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.26.attn2.to_k.weight": { |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.26.attn2.to_out.0.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.26.attn2.to_q.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.26.attn2.to_v.weight": { |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.26.ff.net.0.proj.weight": { |
| "shape": [ |
| 8192, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.26.ff.net.2.weight": { |
| "shape": [ |
| 2048, |
| 8192 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.26.norm1.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.26.norm1.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.26.norm2.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.26.norm2.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.26.norm3.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.26.norm3.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.27.attn1.norm_k.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.27.attn1.norm_q.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.27.attn1.to_k.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.27.attn1.to_out.0.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.27.attn1.to_q.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.27.attn1.to_v.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.27.attn2.norm_k.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.27.attn2.norm_q.weight": { |
| "shape": [ |
| 128 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.27.attn2.to_k.weight": { |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.27.attn2.to_out.0.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.27.attn2.to_q.weight": { |
| "shape": [ |
| 2048, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.27.attn2.to_v.weight": { |
| "shape": [ |
| 2048, |
| 1024 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.27.ff.net.0.proj.weight": { |
| "shape": [ |
| 8192, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.27.ff.net.2.weight": { |
| "shape": [ |
| 2048, |
| 8192 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.27.norm1.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.27.norm1.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.27.norm2.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.27.norm2.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.27.norm3.linear_1.weight": { |
| "shape": [ |
| 256, |
| 2048 |
| ], |
| "dtype": "torch.bfloat16" |
| }, |
| "transformer_blocks.27.norm3.linear_2.weight": { |
| "shape": [ |
| 6144, |
| 256 |
| ], |
| "dtype": "torch.bfloat16" |
| } |
| }, |
| "all_tensors_exact_bf16_cast": true, |
| "k_conditioning_preserved": true, |
| "source_vs_export_bf16_forward_max_abs_errors": [ |
| 0.0, |
| 0.0, |
| 0.0, |
| 0.0 |
| ], |
| "transformer_sha256": "8ea644535ab333540320c6c3d13969c4487c1a6f1135898322317b0b0fb4bf99", |
| "former_hub_revision": "ad1cb61e3d569dc01b748f83ceb58a8199583cee" |
| } |
|
|
