config.json
| 1 | { |
| 2 | "activation_dropout": 0.1, |
| 3 | "apply_spec_augment": true, |
| 4 | "architectures": [ |
| 5 | "SpeechT5ForTextToSpeech" |
| 6 | ], |
| 7 | "attention_dropout": 0.1, |
| 8 | "bos_token_id": 0, |
| 9 | "conv_bias": false, |
| 10 | "conv_dim": [ |
| 11 | 512, |
| 12 | 512, |
| 13 | 512, |
| 14 | 512, |
| 15 | 512, |
| 16 | 512, |
| 17 | 512 |
| 18 | ], |
| 19 | "conv_kernel": [ |
| 20 | 10, |
| 21 | 3, |
| 22 | 3, |
| 23 | 3, |
| 24 | 3, |
| 25 | 2, |
| 26 | 2 |
| 27 | ], |
| 28 | "conv_stride": [ |
| 29 | 5, |
| 30 | 2, |
| 31 | 2, |
| 32 | 2, |
| 33 | 2, |
| 34 | 2, |
| 35 | 2 |
| 36 | ], |
| 37 | "decoder_attention_heads": 12, |
| 38 | "decoder_ffn_dim": 3072, |
| 39 | "decoder_layerdrop": 0.1, |
| 40 | "decoder_layers": 6, |
| 41 | "decoder_start_token_id": 2, |
| 42 | "encoder_attention_heads": 12, |
| 43 | "encoder_ffn_dim": 3072, |
| 44 | "encoder_layerdrop": 0.1, |
| 45 | "encoder_layers": 12, |
| 46 | "encoder_max_relative_position": 160, |
| 47 | "eos_token_id": 2, |
| 48 | "feat_extract_activation": "gelu", |
| 49 | "feat_extract_norm": "group", |
| 50 | "feat_proj_dropout": 0.0, |
| 51 | "guided_attention_loss_num_heads": 2, |
| 52 | "guided_attention_loss_scale": 10.0, |
| 53 | "guided_attention_loss_sigma": 0.4, |
| 54 | "hidden_act": "gelu", |
| 55 | "hidden_dropout": 0.1, |
| 56 | "hidden_size": 768, |
| 57 | "initializer_range": 0.02, |
| 58 | "is_encoder_decoder": true, |
| 59 | "layer_norm_eps": 1e-05, |
| 60 | "mask_feature_length": 10, |
| 61 | "mask_feature_min_masks": 0, |
| 62 | "mask_feature_prob": 0.0, |
| 63 | "mask_time_length": 10, |
| 64 | "mask_time_min_masks": 2, |
| 65 | "mask_time_prob": 0.05, |
| 66 | "max_length": 1876, |
| 67 | "max_speech_positions": 1876, |
| 68 | "max_text_positions": 600, |
| 69 | "model_type": "speecht5", |
| 70 | "num_conv_pos_embedding_groups": 16, |
| 71 | "num_conv_pos_embeddings": 128, |
| 72 | "num_feat_extract_layers": 7, |
| 73 | "num_mel_bins": 80, |
| 74 | "pad_token_id": 1, |
| 75 | "positional_dropout": 0.1, |
| 76 | "reduction_factor": 2, |
| 77 | "scale_embedding": false, |
| 78 | "speaker_embedding_dim": 512, |
| 79 | "speech_decoder_postnet_dropout": 0.5, |
| 80 | "speech_decoder_postnet_kernel": 5, |
| 81 | "speech_decoder_postnet_layers": 5, |
| 82 | "speech_decoder_postnet_units": 256, |
| 83 | "speech_decoder_prenet_dropout": 0.5, |
| 84 | "speech_decoder_prenet_layers": 2, |
| 85 | "speech_decoder_prenet_units": 256, |
| 86 | "torch_dtype": "float32", |
| 87 | "transformers_version": "4.28.0.dev0", |
| 88 | "use_cache": true, |
| 89 | "use_guided_attention_loss": true, |
| 90 | "vocab_size": 81 |
| 91 | } |
| 92 | |