| { |
| "architectures": [ |
| "ChatterboxModel" |
| ], |
| "dtype": "float32", |
| "hiftnet_config": { |
| "audio_limit": 0.99, |
| "base_channels": 512, |
| "f0_predictor_cond_channels": 512, |
| "f0_predictor_in_channels": 80, |
| "hidden_size": 512, |
| "in_channels": 80, |
| "istft_hop_len": 4, |
| "istft_n_fft": 16, |
| "lrelu_slope": 0.1, |
| "model_type": "hiftnet", |
| "nb_harmonics": 8, |
| "nsf_alpha": 0.1, |
| "nsf_sigma": 0.003, |
| "nsf_voiced_threshold": 10.0, |
| "resblock_dilation_sizes": [ |
| [ |
| 1, |
| 3, |
| 5 |
| ], |
| [ |
| 1, |
| 3, |
| 5 |
| ], |
| [ |
| 1, |
| 3, |
| 5 |
| ] |
| ], |
| "resblock_kernel_sizes": [ |
| 3, |
| 7, |
| 11 |
| ], |
| "sampling_rate": 22050, |
| "source_resblock_dilation_sizes": [ |
| [ |
| 1, |
| 3, |
| 5 |
| ], |
| [ |
| 1, |
| 3, |
| 5 |
| ] |
| ], |
| "source_resblock_kernel_sizes": [ |
| 7, |
| 11 |
| ], |
| "upsample_kernel_sizes": [ |
| 16, |
| 16 |
| ], |
| "upsample_rates": [ |
| 8, |
| 8 |
| ] |
| }, |
| "is_multilingual": false, |
| "model_type": "chatterbox", |
| "s3gen_config": { |
| "cfm_inference_cfg_rate": 0.7, |
| "cfm_sigma_min": 1e-06, |
| "cfm_solver": "euler", |
| "cfm_t_scheduler": "cosine", |
| "decoder_act_fn": "gelu", |
| "decoder_attention_head_dim": 64, |
| "decoder_channels": [ |
| 256 |
| ], |
| "decoder_in_channels": 320, |
| "decoder_n_blocks": 4, |
| "decoder_num_heads": 8, |
| "decoder_num_mid_blocks": 12, |
| "decoder_out_channels": 80, |
| "encoder_attention_heads": 8, |
| "encoder_dropout_rate": 0.1, |
| "encoder_linear_units": 2048, |
| "encoder_num_blocks": 6, |
| "encoder_output_size": 512, |
| "fmax": 8000, |
| "fmin": 0, |
| "hop_length": 480, |
| "input_frame_rate": 25, |
| "mel_bins": 80, |
| "model_type": "s3gen", |
| "n_fft": 1920, |
| "pre_lookahead_len": 3, |
| "sampling_rate": 24000, |
| "speaker_embed_dim": 192, |
| "speaker_feat_dim": 80, |
| "token_embed_dim": 512, |
| "token_mel_ratio": 2, |
| "vocab_size": 6561, |
| "win_size": 1920 |
| }, |
| "t3_config": { |
| "alignment_layer_idx": 9, |
| "emotion_adv": true, |
| "encoder_type": "voice_encoder", |
| "hidden_size": 1024, |
| "input_pos_emb": "learned", |
| "llama_config_dict": { |
| "attention_bias": false, |
| "attention_dropout": 0.0, |
| "attn_implementation": "sdpa", |
| "head_dim": 64, |
| "hidden_act": "silu", |
| "hidden_size": 1024, |
| "initializer_range": 0.02, |
| "intermediate_size": 4096, |
| "max_position_embeddings": 131072, |
| "mlp_bias": false, |
| "model_type": "llama", |
| "num_attention_heads": 16, |
| "num_hidden_layers": 30, |
| "num_key_value_heads": 16, |
| "pretraining_tp": 1, |
| "rms_norm_eps": 1e-05, |
| "rope_scaling": { |
| "factor": 8.0, |
| "high_freq_factor": 4.0, |
| "low_freq_factor": 1.0, |
| "original_max_position_embeddings": 8192, |
| "rope_theta": 500000.0, |
| "rope_type": "llama3" |
| }, |
| "rope_theta": 500000.0, |
| "tie_word_embeddings": false, |
| "torch_dtype": "bfloat16", |
| "use_cache": true, |
| "vocab_size": 8 |
| }, |
| "llama_config_name": "Llama_520M", |
| "max_speech_tokens": 4096, |
| "max_text_tokens": 2048, |
| "model_type": "t3", |
| "perceiver_latent_dim": 1024, |
| "perceiver_num_heads": 4, |
| "perceiver_num_latents": 32, |
| "speaker_embed_size": 256, |
| "speech_cond_prompt_len": 150, |
| "speech_tokens_dict_size": 8194, |
| "start_speech_token": 6561, |
| "start_text_token": 255, |
| "stop_speech_token": 6562, |
| "stop_text_token": 0, |
| "text_tokens_dict_size": 704, |
| "use_alignment_analyzer": false, |
| "use_perceiver_resampler": true |
| }, |
| "transformers_version": "5.0.0.dev0" |
| } |
|
|
