| *.7z filter=lfs diff=lfs merge=lfs -text | |
| *.arrow filter=lfs diff=lfs merge=lfs -text | |
| *.bin filter=lfs diff=lfs merge=lfs -text | |
| *.bz2 filter=lfs diff=lfs merge=lfs -text | |
| *.ckpt filter=lfs diff=lfs merge=lfs -text | |
| *.ftz filter=lfs diff=lfs merge=lfs -text | |
| *.gz filter=lfs diff=lfs merge=lfs -text | |
| *.h5 filter=lfs diff=lfs merge=lfs -text | |
| *.joblib filter=lfs diff=lfs merge=lfs -text | |
| *.lfs.* filter=lfs diff=lfs merge=lfs -text | |
| *.mlmodel filter=lfs diff=lfs merge=lfs -text | |
| *.model filter=lfs diff=lfs merge=lfs -text | |
| *.msgpack filter=lfs diff=lfs merge=lfs -text | |
| *.npy filter=lfs diff=lfs merge=lfs -text | |
| *.npz filter=lfs diff=lfs merge=lfs -text | |
| *.onnx filter=lfs diff=lfs merge=lfs -text | |
| *.ot filter=lfs diff=lfs merge=lfs -text | |
| *.parquet filter=lfs diff=lfs merge=lfs -text | |
| *.pb filter=lfs diff=lfs merge=lfs -text | |
| *.pickle filter=lfs diff=lfs merge=lfs -text | |
| *.pkl filter=lfs diff=lfs merge=lfs -text | |
| *.pt filter=lfs diff=lfs merge=lfs -text | |
| *.pth filter=lfs diff=lfs merge=lfs -text | |
| *.rar filter=lfs diff=lfs merge=lfs -text | |
| *.safetensors filter=lfs diff=lfs merge=lfs -text | |
| saved_model/**/* filter=lfs diff=lfs merge=lfs -text | |
| *.tar.* filter=lfs diff=lfs merge=lfs -text | |
| *.tar filter=lfs diff=lfs merge=lfs -text | |
| *.tflite filter=lfs diff=lfs merge=lfs -text | |
| *.tgz filter=lfs diff=lfs merge=lfs -text | |
| *.wasm filter=lfs diff=lfs merge=lfs -text | |
| *.xz filter=lfs diff=lfs merge=lfs -text | |
| *.zip filter=lfs diff=lfs merge=lfs -text | |
| *.zst filter=lfs diff=lfs merge=lfs -text | |
| *tfevents* filter=lfs diff=lfs merge=lfs -text | |
| decoder.embed_positions.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.embed_tokens.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.-1.attn-head-sparsity-predictor.1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.0.attn-head-sparsity-predictor.1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.0.attn.mlp-sparsity-predictor.1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.0.attn.mlp-sparsity-predictor.2.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.0.fc1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.0.fc2.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.0.self_attn.catted_head_weights filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.1.attn-head-sparsity-predictor.1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.1.attn.mlp-sparsity-predictor.1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.1.attn.mlp-sparsity-predictor.2.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.1.fc1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.1.fc2.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.1.self_attn.catted_head_weights filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.10.attn-head-sparsity-predictor.1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.10.attn.mlp-sparsity-predictor.1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.10.attn.mlp-sparsity-predictor.2.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.10.fc1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.10.fc2.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.10.self_attn.catted_head_weights filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.11.attn-head-sparsity-predictor.1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.11.attn.mlp-sparsity-predictor.1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.11.attn.mlp-sparsity-predictor.2.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.11.fc1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.11.fc2.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.11.self_attn.catted_head_weights filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.12.attn-head-sparsity-predictor.1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.12.attn.mlp-sparsity-predictor.1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.12.attn.mlp-sparsity-predictor.2.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.12.fc1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.12.fc2.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.12.self_attn.catted_head_weights filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.13.attn-head-sparsity-predictor.1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.13.attn.mlp-sparsity-predictor.1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.13.attn.mlp-sparsity-predictor.2.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.13.fc1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.13.fc2.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.13.self_attn.catted_head_weights filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.14.attn-head-sparsity-predictor.1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.14.attn.mlp-sparsity-predictor.1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.14.attn.mlp-sparsity-predictor.2.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.14.fc1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.14.fc2.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.14.self_attn.catted_head_weights filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.15.attn-head-sparsity-predictor.1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.15.attn.mlp-sparsity-predictor.1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.15.attn.mlp-sparsity-predictor.2.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.15.fc1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.15.fc2.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.15.self_attn.catted_head_weights filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.16.attn-head-sparsity-predictor.1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.16.attn.mlp-sparsity-predictor.1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.16.attn.mlp-sparsity-predictor.2.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.16.fc1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.16.fc2.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.16.self_attn.catted_head_weights filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.17.attn-head-sparsity-predictor.1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.17.attn.mlp-sparsity-predictor.1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.17.attn.mlp-sparsity-predictor.2.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.17.fc1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.17.fc2.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.17.self_attn.catted_head_weights filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.18.attn-head-sparsity-predictor.1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.18.attn.mlp-sparsity-predictor.1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.18.attn.mlp-sparsity-predictor.2.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.18.fc1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.18.fc2.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.18.self_attn.catted_head_weights filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.19.attn-head-sparsity-predictor.1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.19.attn.mlp-sparsity-predictor.1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.19.attn.mlp-sparsity-predictor.2.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.19.fc1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.19.fc2.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.19.self_attn.catted_head_weights filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.2.attn-head-sparsity-predictor.1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.2.attn.mlp-sparsity-predictor.1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.2.attn.mlp-sparsity-predictor.2.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.2.fc1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.2.fc2.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.2.self_attn.catted_head_weights filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.20.attn-head-sparsity-predictor.1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.20.attn.mlp-sparsity-predictor.1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.20.attn.mlp-sparsity-predictor.2.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.20.fc1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.20.fc2.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.20.self_attn.catted_head_weights filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.21.attn-head-sparsity-predictor.1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.21.attn.mlp-sparsity-predictor.1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.21.attn.mlp-sparsity-predictor.2.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.21.fc1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.21.fc2.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.21.self_attn.catted_head_weights filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.22.attn-head-sparsity-predictor.1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.22.attn.mlp-sparsity-predictor.1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.22.attn.mlp-sparsity-predictor.2.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.22.fc1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.22.fc2.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.22.self_attn.catted_head_weights filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.23.attn-head-sparsity-predictor.1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.23.attn.mlp-sparsity-predictor.1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.23.attn.mlp-sparsity-predictor.2.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.23.fc1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.23.fc2.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.23.self_attn.catted_head_weights filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.24.attn-head-sparsity-predictor.1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.24.attn.mlp-sparsity-predictor.1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.24.attn.mlp-sparsity-predictor.2.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.24.fc1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.24.fc2.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.24.self_attn.catted_head_weights filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.25.attn-head-sparsity-predictor.1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.25.attn.mlp-sparsity-predictor.1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.25.attn.mlp-sparsity-predictor.2.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.25.fc1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.25.fc2.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.25.self_attn.catted_head_weights filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.26.attn-head-sparsity-predictor.1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.26.attn.mlp-sparsity-predictor.1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.26.attn.mlp-sparsity-predictor.2.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.26.fc1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.26.fc2.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.26.self_attn.catted_head_weights filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.27.attn-head-sparsity-predictor.1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.27.attn.mlp-sparsity-predictor.1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.27.attn.mlp-sparsity-predictor.2.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.27.fc1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.27.fc2.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.27.self_attn.catted_head_weights filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.28.attn-head-sparsity-predictor.1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.28.attn.mlp-sparsity-predictor.1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.28.attn.mlp-sparsity-predictor.2.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.28.fc1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.28.fc2.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.28.self_attn.catted_head_weights filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.29.attn-head-sparsity-predictor.1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.29.attn.mlp-sparsity-predictor.1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.29.attn.mlp-sparsity-predictor.2.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.29.fc1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.29.fc2.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.29.self_attn.catted_head_weights filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.3.attn-head-sparsity-predictor.1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.3.attn.mlp-sparsity-predictor.1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.3.attn.mlp-sparsity-predictor.2.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.3.fc1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.3.fc2.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.3.self_attn.catted_head_weights filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.30.attn-head-sparsity-predictor.1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.30.attn.mlp-sparsity-predictor.1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.30.attn.mlp-sparsity-predictor.2.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.30.fc1.weight filter=lfs diff=lfs merge=lfs -text | |
| decoder.layers.30.fc2.weight filter=lfs diff=lfs merge=lfs -text | |