| { | |
| "model_type": "aura", | |
| "base_model": "Qwen/Qwen3-8B", | |
| "codec": "X-Codec xcodec_hubert_general_audio_v2 (frozen), 8 RVQ layers, 16 kHz, 50 Hz", | |
| "slots": [ | |
| "left_content", | |
| "left_sound", | |
| "right_content", | |
| "right_sound", | |
| "side_content", | |
| "side_sound" | |
| ], | |
| "side_definition": "S=(L-R)/2 after codebook lookup, before pooling", | |
| "content_cut": 3, | |
| "time_pool": 4, | |
| "pool_content": "max", | |
| "pool_sound": "mean", | |
| "audio_frames": 500, | |
| "projector": { | |
| "type": "mlp_gelu", | |
| "in_dim": 1024, | |
| "hidden_dim": 4096, | |
| "out_dim": 4096 | |
| }, | |
| "lora": { | |
| "r": 16, | |
| "alpha": 32, | |
| "dropout": 0.05, | |
| "target_modules": [ | |
| "q_proj", | |
| "k_proj", | |
| "v_proj", | |
| "o_proj", | |
| "gate_proj", | |
| "up_proj", | |
| "down_proj" | |
| ] | |
| } | |
| } | |