config.json
| 1 | { |
| 2 | "_name_or_path": "m-a-p/MERT-v1-330M", |
| 3 | "activation_dropout": 0.0, |
| 4 | "apply_spec_augment": true, |
| 5 | "architectures": [ |
| 6 | "MERTModel" |
| 7 | ], |
| 8 | "attention_dropout": 0.0, |
| 9 | "attention_relax": 32.0, |
| 10 | "auto_map": { |
| 11 | "AutoConfig": "configuration_MERT.MERTConfig", |
| 12 | "AutoModel": "modeling_MERT.MERTModel" |
| 13 | }, |
| 14 | "bos_token_id": 1, |
| 15 | "classifier_proj_size": 256, |
| 16 | "conv_bias": false, |
| 17 | "conv_dim": [ |
| 18 | 512, |
| 19 | 512, |
| 20 | 512, |
| 21 | 512, |
| 22 | 512, |
| 23 | 512, |
| 24 | 512 |
| 25 | ], |
| 26 | "conv_kernel": [ |
| 27 | 10, |
| 28 | 3, |
| 29 | 3, |
| 30 | 3, |
| 31 | 3, |
| 32 | 2, |
| 33 | 2 |
| 34 | ], |
| 35 | "conv_stride": [ |
| 36 | 5, |
| 37 | 2, |
| 38 | 2, |
| 39 | 2, |
| 40 | 2, |
| 41 | 2, |
| 42 | 2 |
| 43 | ], |
| 44 | "ctc_loss_reduction": "sum", |
| 45 | "ctc_zero_infinity": false, |
| 46 | "deepnorm": false, |
| 47 | "do_stable_layer_norm": true, |
| 48 | "eos_token_id": 2, |
| 49 | "feat_extract_activation": "gelu", |
| 50 | "feat_extract_dropout": 0.0, |
| 51 | "feat_extract_norm": "group", |
| 52 | "feat_proj_dropout": 0.0, |
| 53 | "feat_proj_layer_norm": true, |
| 54 | "feature_extractor_cqt": false, |
| 55 | "feature_extractor_cqt_bins": 336, |
| 56 | "final_dropout": 0.0, |
| 57 | "gradient_checkpointing": false, |
| 58 | "hidden_act": "gelu", |
| 59 | "hidden_dropout": 0.0, |
| 60 | "hidden_size": 1024, |
| 61 | "initializer_range": 0.02, |
| 62 | "intermediate_size": 4096, |
| 63 | "layer_norm_eps": 1e-05, |
| 64 | "layerdrop": 0.0, |
| 65 | "mask_channel_length": 10, |
| 66 | "mask_channel_min_space": 1, |
| 67 | "mask_channel_other": 0.0, |
| 68 | "mask_channel_prob": 0.0, |
| 69 | "mask_channel_selection": "static", |
| 70 | "mask_feature_length": 10, |
| 71 | "mask_feature_min_masks": 0, |
| 72 | "mask_feature_prob": 0.0, |
| 73 | "mask_time_length": 10, |
| 74 | "mask_time_min_masks": 2, |
| 75 | "mask_time_min_space": 1, |
| 76 | "mask_time_other": 0.0, |
| 77 | "mask_time_prob": 0.075, |
| 78 | "mask_time_selection": "static", |
| 79 | "model_type": "mert_model", |
| 80 | "num_attention_heads": 16, |
| 81 | "num_conv_pos_embedding_groups": 16, |
| 82 | "num_conv_pos_embeddings": 128, |
| 83 | "num_feat_extract_layers": 7, |
| 84 | "num_hidden_layers": 24, |
| 85 | "pad_token_id": 0, |
| 86 | "sample_rate": 24000, |
| 87 | "tokenizer_class": "Wav2Vec2CTCTokenizer", |
| 88 | "torch_dtype": "float32", |
| 89 | "transformers_version": "4.27.1", |
| 90 | "use_weighted_layer_sum": false, |
| 91 | "vocab_size": 32 |
| 92 | } |
| 93 | |