| { |
| "checkpoint_id": "Duo-1.2-causal-base", |
| "files": { |
| "dcp/.metadata": { |
| "sha256": "1022e7278d4a062bba00c70289b5c08bb115be4689ff2d55c010813a09ef9c1c", |
| "size": 206415 |
| }, |
| "dcp/__0_0.distcp": { |
| "sha256": "05332963e83056c3b3352329fdd55e84fb46881170f7889ddebf7ce7eede88a7", |
| "size": 2784751539 |
| } |
| }, |
| "format": "duo-release-dcp-v1", |
| "identity": { |
| "codec": { |
| "audio_hop": 800, |
| "audio_latents_per_block": 29, |
| "audio_sample_rate": 32000, |
| "decode_policy": "exact_delay_eof_last_two", |
| "feature_dtype": "torch.bfloat16", |
| "fps": 24, |
| "frame_spans": [ |
| 1, |
| 4, |
| 4, |
| 4, |
| 4 |
| ], |
| "frames_per_block": 17, |
| "mode": "duo_5_17", |
| "normalization": "duo_channel_mean_std", |
| "posterior": "mean", |
| "vae_dtype": "torch.float32", |
| "version": 1, |
| "video_autocast_dtype": "torch.bfloat16", |
| "video_latents_per_block": 5, |
| "weights": { |
| "diffusion_pytorch_model.safetensors": "52c59e67ba8de5477c81bfbced0327aabf500f1bfdeefd5ee754529241cb26cb", |
| "model.safetensors": "5f0c2e161d895a9fee7645ca32d4a7e3a22b90cacfcbeba62ec999cdbbefe0d3" |
| } |
| }, |
| "intent": { |
| "adapter_weights": { |
| "adapter_model.safetensors": "2f7de1b18f515398f3c56c35b7472c9a15033ddd7f544822c6786bf91853d021" |
| }, |
| "chunk_token": "<chunk_708>", |
| "fps": 24, |
| "frames_per_block": 17, |
| "protocol_manifest_sha256": "bb650b4233bb62781208bb3e59665111455a6f6ebdcc4f0390b4ebe8729c48db" |
| }, |
| "realization": { |
| "condition_dropout": 0.0, |
| "mode": "lora", |
| "role_embedding": false |
| } |
| }, |
| "inference": { |
| "flow_shift": 1.0, |
| "sampling_steps": 50 |
| }, |
| "models": { |
| "student": { |
| "base": { |
| "config_sha256": "f619093a231fcfbcc3d035bec26c50ad864e7331a500d5c519f5045dc1e50458", |
| "format": "original_duo", |
| "index_sha256": "fb457a26ffa6294660e249b0ddd03a337f2e5393f770b5c34c8b8f90a29a7efb", |
| "path": "backbone" |
| }, |
| "config": { |
| "attention_head_dim": 128, |
| "audio_in_channels": 32, |
| "ffn_dim": 14336, |
| "final_norm_eps": 1e-05, |
| "freq_dim": 256, |
| "hidden_size": 5376, |
| "in_channels": 24, |
| "norm_eps": 1e-05, |
| "num_attention_heads": 56, |
| "num_layers": 50, |
| "num_refiner_layers": 2, |
| "patch_size": [ |
| 1, |
| 2, |
| 2 |
| ], |
| "qk_norm_eps": 1e-05, |
| "rope_freq_dim": 16, |
| "rope_theta": 10000.0, |
| "text_dim": 5120, |
| "time_embed_dim": 2688, |
| "time_embed_hidden_dim": 5376 |
| }, |
| "decfg": { |
| "base_model": "ref2va", |
| "conversion_version": 1, |
| "enabled": true, |
| "rank": 64, |
| "scale": 1.0, |
| "sha256": "9361683349c1bf78f206c382d91b1e01b91fcddc2408e85ab39c9faa51f0d3d1" |
| }, |
| "lora": { |
| "alpha": 128, |
| "modules": [ |
| "audio_proj_in", |
| "audio_proj_out", |
| "context_embedder", |
| "proj_in", |
| "proj_out", |
| "token_refiner.refiner_blocks.0.attn.to_k", |
| "token_refiner.refiner_blocks.0.attn.to_out.0", |
| "token_refiner.refiner_blocks.0.attn.to_q", |
| "token_refiner.refiner_blocks.0.attn.to_v", |
| "token_refiner.refiner_blocks.0.ff.net.0.proj", |
| "token_refiner.refiner_blocks.0.ff.net.2", |
| "token_refiner.refiner_blocks.1.attn.to_k", |
| "token_refiner.refiner_blocks.1.attn.to_out.0", |
| "token_refiner.refiner_blocks.1.attn.to_q", |
| "token_refiner.refiner_blocks.1.attn.to_v", |
| "token_refiner.refiner_blocks.1.ff.net.0.proj", |
| "token_refiner.refiner_blocks.1.ff.net.2", |
| "transformer_blocks.0.attn.to_k", |
| "transformer_blocks.0.attn.to_out.0", |
| "transformer_blocks.0.attn.to_q", |
| "transformer_blocks.0.attn.to_v", |
| "transformer_blocks.0.ff.net.0.proj", |
| "transformer_blocks.0.ff.net.2", |
| "transformer_blocks.1.attn.to_k", |
| "transformer_blocks.1.attn.to_out.0", |
| "transformer_blocks.1.attn.to_q", |
| "transformer_blocks.1.attn.to_v", |
| "transformer_blocks.1.ff.net.0.proj", |
| "transformer_blocks.1.ff.net.2", |
| "transformer_blocks.10.attn.to_k", |
| "transformer_blocks.10.attn.to_out.0", |
| "transformer_blocks.10.attn.to_q", |
| "transformer_blocks.10.attn.to_v", |
| "transformer_blocks.10.ff.net.0.proj", |
| "transformer_blocks.10.ff.net.2", |
| "transformer_blocks.11.attn.to_k", |
| "transformer_blocks.11.attn.to_out.0", |
| "transformer_blocks.11.attn.to_q", |
| "transformer_blocks.11.attn.to_v", |
| "transformer_blocks.11.ff.net.0.proj", |
| "transformer_blocks.11.ff.net.2", |
| "transformer_blocks.12.attn.to_k", |
| "transformer_blocks.12.attn.to_out.0", |
| "transformer_blocks.12.attn.to_q", |
| "transformer_blocks.12.attn.to_v", |
| "transformer_blocks.12.ff.net.0.proj", |
| "transformer_blocks.12.ff.net.2", |
| "transformer_blocks.13.attn.to_k", |
| "transformer_blocks.13.attn.to_out.0", |
| "transformer_blocks.13.attn.to_q", |
| "transformer_blocks.13.attn.to_v", |
| "transformer_blocks.13.ff.net.0.proj", |
| "transformer_blocks.13.ff.net.2", |
| "transformer_blocks.14.attn.to_k", |
| "transformer_blocks.14.attn.to_out.0", |
| "transformer_blocks.14.attn.to_q", |
| "transformer_blocks.14.attn.to_v", |
| "transformer_blocks.14.ff.net.0.proj", |
| "transformer_blocks.14.ff.net.2", |
| "transformer_blocks.15.attn.to_k", |
| "transformer_blocks.15.attn.to_out.0", |
| "transformer_blocks.15.attn.to_q", |
| "transformer_blocks.15.attn.to_v", |
| "transformer_blocks.15.ff.net.0.proj", |
| "transformer_blocks.15.ff.net.2", |
| "transformer_blocks.16.attn.to_k", |
| "transformer_blocks.16.attn.to_out.0", |
| "transformer_blocks.16.attn.to_q", |
| "transformer_blocks.16.attn.to_v", |
| "transformer_blocks.16.ff.net.0.proj", |
| "transformer_blocks.16.ff.net.2", |
| "transformer_blocks.17.attn.to_k", |
| "transformer_blocks.17.attn.to_out.0", |
| "transformer_blocks.17.attn.to_q", |
| "transformer_blocks.17.attn.to_v", |
| "transformer_blocks.17.ff.net.0.proj", |
| "transformer_blocks.17.ff.net.2", |
| "transformer_blocks.18.attn.to_k", |
| "transformer_blocks.18.attn.to_out.0", |
| "transformer_blocks.18.attn.to_q", |
| "transformer_blocks.18.attn.to_v", |
| "transformer_blocks.18.ff.net.0.proj", |
| "transformer_blocks.18.ff.net.2", |
| "transformer_blocks.19.attn.to_k", |
| "transformer_blocks.19.attn.to_out.0", |
| "transformer_blocks.19.attn.to_q", |
| "transformer_blocks.19.attn.to_v", |
| "transformer_blocks.19.ff.net.0.proj", |
| "transformer_blocks.19.ff.net.2", |
| "transformer_blocks.2.attn.to_k", |
| "transformer_blocks.2.attn.to_out.0", |
| "transformer_blocks.2.attn.to_q", |
| "transformer_blocks.2.attn.to_v", |
| "transformer_blocks.2.ff.net.0.proj", |
| "transformer_blocks.2.ff.net.2", |
| "transformer_blocks.20.attn.to_k", |
| "transformer_blocks.20.attn.to_out.0", |
| "transformer_blocks.20.attn.to_q", |
| "transformer_blocks.20.attn.to_v", |
| "transformer_blocks.20.ff.net.0.proj", |
| "transformer_blocks.20.ff.net.2", |
| "transformer_blocks.21.attn.to_k", |
| "transformer_blocks.21.attn.to_out.0", |
| "transformer_blocks.21.attn.to_q", |
| "transformer_blocks.21.attn.to_v", |
| "transformer_blocks.21.ff.net.0.proj", |
| "transformer_blocks.21.ff.net.2", |
| "transformer_blocks.22.attn.to_k", |
| "transformer_blocks.22.attn.to_out.0", |
| "transformer_blocks.22.attn.to_q", |
| "transformer_blocks.22.attn.to_v", |
| "transformer_blocks.22.ff.net.0.proj", |
| "transformer_blocks.22.ff.net.2", |
| "transformer_blocks.23.attn.to_k", |
| "transformer_blocks.23.attn.to_out.0", |
| "transformer_blocks.23.attn.to_q", |
| "transformer_blocks.23.attn.to_v", |
| "transformer_blocks.23.ff.net.0.proj", |
| "transformer_blocks.23.ff.net.2", |
| "transformer_blocks.24.attn.to_k", |
| "transformer_blocks.24.attn.to_out.0", |
| "transformer_blocks.24.attn.to_q", |
| "transformer_blocks.24.attn.to_v", |
| "transformer_blocks.24.ff.net.0.proj", |
| "transformer_blocks.24.ff.net.2", |
| "transformer_blocks.25.attn.to_k", |
| "transformer_blocks.25.attn.to_out.0", |
| "transformer_blocks.25.attn.to_q", |
| "transformer_blocks.25.attn.to_v", |
| "transformer_blocks.25.ff.net.0.proj", |
| "transformer_blocks.25.ff.net.2", |
| "transformer_blocks.26.attn.to_k", |
| "transformer_blocks.26.attn.to_out.0", |
| "transformer_blocks.26.attn.to_q", |
| "transformer_blocks.26.attn.to_v", |
| "transformer_blocks.26.ff.net.0.proj", |
| "transformer_blocks.26.ff.net.2", |
| "transformer_blocks.27.attn.to_k", |
| "transformer_blocks.27.attn.to_out.0", |
| "transformer_blocks.27.attn.to_q", |
| "transformer_blocks.27.attn.to_v", |
| "transformer_blocks.27.ff.net.0.proj", |
| "transformer_blocks.27.ff.net.2", |
| "transformer_blocks.28.attn.to_k", |
| "transformer_blocks.28.attn.to_out.0", |
| "transformer_blocks.28.attn.to_q", |
| "transformer_blocks.28.attn.to_v", |
| "transformer_blocks.28.ff.net.0.proj", |
| "transformer_blocks.28.ff.net.2", |
| "transformer_blocks.29.attn.to_k", |
| "transformer_blocks.29.attn.to_out.0", |
| "transformer_blocks.29.attn.to_q", |
| "transformer_blocks.29.attn.to_v", |
| "transformer_blocks.29.ff.net.0.proj", |
| "transformer_blocks.29.ff.net.2", |
| "transformer_blocks.3.attn.to_k", |
| "transformer_blocks.3.attn.to_out.0", |
| "transformer_blocks.3.attn.to_q", |
| "transformer_blocks.3.attn.to_v", |
| "transformer_blocks.3.ff.net.0.proj", |
| "transformer_blocks.3.ff.net.2", |
| "transformer_blocks.30.attn.to_k", |
| "transformer_blocks.30.attn.to_out.0", |
| "transformer_blocks.30.attn.to_q", |
| "transformer_blocks.30.attn.to_v", |
| "transformer_blocks.30.ff.net.0.proj", |
| "transformer_blocks.30.ff.net.2", |
| "transformer_blocks.31.attn.to_k", |
| "transformer_blocks.31.attn.to_out.0", |
| "transformer_blocks.31.attn.to_q", |
| "transformer_blocks.31.attn.to_v", |
| "transformer_blocks.31.ff.net.0.proj", |
| "transformer_blocks.31.ff.net.2", |
| "transformer_blocks.32.attn.to_k", |
| "transformer_blocks.32.attn.to_out.0", |
| "transformer_blocks.32.attn.to_q", |
| "transformer_blocks.32.attn.to_v", |
| "transformer_blocks.32.ff.net.0.proj", |
| "transformer_blocks.32.ff.net.2", |
| "transformer_blocks.33.attn.to_k", |
| "transformer_blocks.33.attn.to_out.0", |
| "transformer_blocks.33.attn.to_q", |
| "transformer_blocks.33.attn.to_v", |
| "transformer_blocks.33.ff.net.0.proj", |
| "transformer_blocks.33.ff.net.2", |
| "transformer_blocks.34.attn.to_k", |
| "transformer_blocks.34.attn.to_out.0", |
| "transformer_blocks.34.attn.to_q", |
| "transformer_blocks.34.attn.to_v", |
| "transformer_blocks.34.ff.net.0.proj", |
| "transformer_blocks.34.ff.net.2", |
| "transformer_blocks.35.attn.to_k", |
| "transformer_blocks.35.attn.to_out.0", |
| "transformer_blocks.35.attn.to_q", |
| "transformer_blocks.35.attn.to_v", |
| "transformer_blocks.35.ff.net.0.proj", |
| "transformer_blocks.35.ff.net.2", |
| "transformer_blocks.36.attn.to_k", |
| "transformer_blocks.36.attn.to_out.0", |
| "transformer_blocks.36.attn.to_q", |
| "transformer_blocks.36.attn.to_v", |
| "transformer_blocks.36.ff.net.0.proj", |
| "transformer_blocks.36.ff.net.2", |
| "transformer_blocks.37.attn.to_k", |
| "transformer_blocks.37.attn.to_out.0", |
| "transformer_blocks.37.attn.to_q", |
| "transformer_blocks.37.attn.to_v", |
| "transformer_blocks.37.ff.net.0.proj", |
| "transformer_blocks.37.ff.net.2", |
| "transformer_blocks.38.attn.to_k", |
| "transformer_blocks.38.attn.to_out.0", |
| "transformer_blocks.38.attn.to_q", |
| "transformer_blocks.38.attn.to_v", |
| "transformer_blocks.38.ff.net.0.proj", |
| "transformer_blocks.38.ff.net.2", |
| "transformer_blocks.39.attn.to_k", |
| "transformer_blocks.39.attn.to_out.0", |
| "transformer_blocks.39.attn.to_q", |
| "transformer_blocks.39.attn.to_v", |
| "transformer_blocks.39.ff.net.0.proj", |
| "transformer_blocks.39.ff.net.2", |
| "transformer_blocks.4.attn.to_k", |
| "transformer_blocks.4.attn.to_out.0", |
| "transformer_blocks.4.attn.to_q", |
| "transformer_blocks.4.attn.to_v", |
| "transformer_blocks.4.ff.net.0.proj", |
| "transformer_blocks.4.ff.net.2", |
| "transformer_blocks.40.attn.to_k", |
| "transformer_blocks.40.attn.to_out.0", |
| "transformer_blocks.40.attn.to_q", |
| "transformer_blocks.40.attn.to_v", |
| "transformer_blocks.40.ff.net.0.proj", |
| "transformer_blocks.40.ff.net.2", |
| "transformer_blocks.41.attn.to_k", |
| "transformer_blocks.41.attn.to_out.0", |
| "transformer_blocks.41.attn.to_q", |
| "transformer_blocks.41.attn.to_v", |
| "transformer_blocks.41.ff.net.0.proj", |
| "transformer_blocks.41.ff.net.2", |
| "transformer_blocks.42.attn.to_k", |
| "transformer_blocks.42.attn.to_out.0", |
| "transformer_blocks.42.attn.to_q", |
| "transformer_blocks.42.attn.to_v", |
| "transformer_blocks.42.ff.net.0.proj", |
| "transformer_blocks.42.ff.net.2", |
| "transformer_blocks.43.attn.to_k", |
| "transformer_blocks.43.attn.to_out.0", |
| "transformer_blocks.43.attn.to_q", |
| "transformer_blocks.43.attn.to_v", |
| "transformer_blocks.43.ff.net.0.proj", |
| "transformer_blocks.43.ff.net.2", |
| "transformer_blocks.44.attn.to_k", |
| "transformer_blocks.44.attn.to_out.0", |
| "transformer_blocks.44.attn.to_q", |
| "transformer_blocks.44.attn.to_v", |
| "transformer_blocks.44.ff.net.0.proj", |
| "transformer_blocks.44.ff.net.2", |
| "transformer_blocks.45.attn.to_k", |
| "transformer_blocks.45.attn.to_out.0", |
| "transformer_blocks.45.attn.to_q", |
| "transformer_blocks.45.attn.to_v", |
| "transformer_blocks.45.ff.net.0.proj", |
| "transformer_blocks.45.ff.net.2", |
| "transformer_blocks.46.attn.to_k", |
| "transformer_blocks.46.attn.to_out.0", |
| "transformer_blocks.46.attn.to_q", |
| "transformer_blocks.46.attn.to_v", |
| "transformer_blocks.46.ff.net.0.proj", |
| "transformer_blocks.46.ff.net.2", |
| "transformer_blocks.47.attn.to_k", |
| "transformer_blocks.47.attn.to_out.0", |
| "transformer_blocks.47.attn.to_q", |
| "transformer_blocks.47.attn.to_v", |
| "transformer_blocks.47.ff.net.0.proj", |
| "transformer_blocks.47.ff.net.2", |
| "transformer_blocks.48.attn.to_k", |
| "transformer_blocks.48.attn.to_out.0", |
| "transformer_blocks.48.attn.to_q", |
| "transformer_blocks.48.attn.to_v", |
| "transformer_blocks.48.ff.net.0.proj", |
| "transformer_blocks.48.ff.net.2", |
| "transformer_blocks.49.attn.to_k", |
| "transformer_blocks.49.attn.to_out.0", |
| "transformer_blocks.49.attn.to_q", |
| "transformer_blocks.49.attn.to_v", |
| "transformer_blocks.49.ff.net.0.proj", |
| "transformer_blocks.49.ff.net.2", |
| "transformer_blocks.5.attn.to_k", |
| "transformer_blocks.5.attn.to_out.0", |
| "transformer_blocks.5.attn.to_q", |
| "transformer_blocks.5.attn.to_v", |
| "transformer_blocks.5.ff.net.0.proj", |
| "transformer_blocks.5.ff.net.2", |
| "transformer_blocks.6.attn.to_k", |
| "transformer_blocks.6.attn.to_out.0", |
| "transformer_blocks.6.attn.to_q", |
| "transformer_blocks.6.attn.to_v", |
| "transformer_blocks.6.ff.net.0.proj", |
| "transformer_blocks.6.ff.net.2", |
| "transformer_blocks.7.attn.to_k", |
| "transformer_blocks.7.attn.to_out.0", |
| "transformer_blocks.7.attn.to_q", |
| "transformer_blocks.7.attn.to_v", |
| "transformer_blocks.7.ff.net.0.proj", |
| "transformer_blocks.7.ff.net.2", |
| "transformer_blocks.8.attn.to_k", |
| "transformer_blocks.8.attn.to_out.0", |
| "transformer_blocks.8.attn.to_q", |
| "transformer_blocks.8.attn.to_v", |
| "transformer_blocks.8.ff.net.0.proj", |
| "transformer_blocks.8.ff.net.2", |
| "transformer_blocks.9.attn.to_k", |
| "transformer_blocks.9.attn.to_out.0", |
| "transformer_blocks.9.attn.to_q", |
| "transformer_blocks.9.attn.to_v", |
| "transformer_blocks.9.ff.net.0.proj", |
| "transformer_blocks.9.ff.net.2" |
| ], |
| "rank": 128 |
| }, |
| "mode": "lora", |
| "role_embedding": false, |
| "storage": { |
| "format": "dcp" |
| } |
| } |
| }, |
| "release": { |
| "name": "Duo-1.2-causal-base", |
| "realization_storage": "dcp" |
| }, |
| "stage": "causal" |
| } |
|
|
