| { |
| "model_type": "cutetts", |
| "variant": "distill", |
| "sample_rate": 24000, |
| "architecture": { |
| "lm_config": { |
| "architectures": [ |
| "Qwen3ForCausalLM" |
| ], |
| "attention_bias": false, |
| "attention_dropout": 0.0, |
| "bos_token_id": 151643, |
| "eos_token_id": 151643, |
| "head_dim": 128, |
| "hidden_act": "silu", |
| "hidden_size": 1024, |
| "initializer_range": 0.02, |
| "intermediate_size": 3072, |
| "max_position_embeddings": 32768, |
| "max_window_layers": 28, |
| "model_type": "qwen3", |
| "num_attention_heads": 16, |
| "num_hidden_layers": 28, |
| "num_key_value_heads": 8, |
| "rms_norm_eps": 1e-06, |
| "rope_scaling": null, |
| "rope_theta": 1000000, |
| "sliding_window": null, |
| "tie_word_embeddings": true, |
| "torch_dtype": "bfloat16", |
| "use_cache": true, |
| "use_sliding_window": false, |
| "vocab_size": 151936 |
| }, |
| "attn_implementation": "sdpa", |
| "torch_dtype": "bfloat16", |
| "lm_keep_num_hidden_layers": 7, |
| "acoustic_latent_dim": 64, |
| "include_semantic_latent": false, |
| "extended_vocab_size": 16385, |
| "locenc_enabled": true, |
| "locenc_patch_size": 2, |
| "locenc_use_rope": true, |
| "locenc_rope_theta": 10000.0, |
| "locenc_layers": 2, |
| "locenc_hidden_dim": 1024, |
| "locenc_ffn_dim": 4096, |
| "locenc_num_heads": 16, |
| "locenc_num_kv_heads": 2, |
| "diff_head_kind": "audio_dit", |
| "diff_dit_patch_size": 2, |
| "diff_dit_use_rope": true, |
| "diff_dit_rope_theta": 10000.0, |
| "diff_dit_layers": 4, |
| "diff_dit_hidden_dim": 1024, |
| "diff_dit_ffn_dim": 4096, |
| "diff_dit_num_heads": 16, |
| "diff_dit_num_kv_heads": 2, |
| "diff_dit_use_previous_cond": true, |
| "diff_dit_speaker_adaln_zero_enabled": true, |
| "diff_dit_speaker_embedding_dim": 256, |
| "lm_speaker_linear_enabled": true, |
| "lm_speaker_embedding_dim": 256, |
| "diff_latent_history": false, |
| "diff_latent_history_size": 0, |
| "scale_acoustic_latent": true, |
| "diff_dit_cfg_strength_embedding_enabled": true, |
| "diff_dit_cfg_strength_max": 5.0, |
| "diff_dit_step_size_embedding_enabled": true, |
| "diff_dit_step_schedule": "uniform", |
| "two_class_stop_predictor": true |
| }, |
| "processor": { |
| "speech_compress_rate": 1920, |
| "segment": { |
| "max_length": 8192, |
| "batch_pad_token_id": 0, |
| "audio_pad_token_id": 0, |
| "speech_pad_value": 0.0, |
| "audio_feat_dim": 64, |
| "audio_patch_size": 2, |
| "semantic_feat_dim": null, |
| "enable_semantic_features": false, |
| "padding_side": "left" |
| } |
| }, |
| "inference_defaults": { |
| "cfg_strength": 2.0, |
| "diffusion_steps": 4, |
| "diffusion_sway_coefficient": null, |
| "max_decode_length": 750 |
| } |
| } |
|
|
