| { |
| "_name_or_path": "BUT-FIT/DiCoW_v3_MLC", |
| "activation_dropout": 0.0, |
| "activation_function": "gelu", |
| "additional_layer": false, |
| "additional_self_attention_layer": true, |
| "apply_fddt_to_n_layers": -1, |
| "apply_spec_augment": false, |
| "architectures": [ |
| "DiCoWForConditionalGeneration" |
| ], |
| "attention_dropout": 0.0, |
| "auto_map": { |
| "AutoConfig": "config.DiCoWConfig", |
| "AutoModelForSpeechSeq2Seq": "modeling_dicow.DiCoWForConditionalGeneration" |
| }, |
| "begin_suppress_tokens": [ |
| 220, |
| 50256 |
| ], |
| "blank_token_id": null, |
| "bos_token_id": 50257, |
| "classifier_proj_size": 256, |
| "ctc_loss_reduction": "mean", |
| "ctc_weight": 0.3, |
| "ctc_zero_infinity": false, |
| "d_model": 1280, |
| "decoder_attention_heads": 20, |
| "decoder_ffn_dim": 5120, |
| "decoder_layerdrop": 0.0, |
| "decoder_layers": 4, |
| "decoder_start_token_id": 50258, |
| "dropout": 0.0, |
| "encoder_attention_heads": 20, |
| "encoder_ffn_dim": 5120, |
| "encoder_layerdrop": 0.0, |
| "encoder_layers": 32, |
| "eos_token_id": 50257, |
| "fddt_bias_only": false, |
| "fddt_init": "disparagement", |
| "fddt_is_diagonal": true, |
| "fddt_use_non_target": true, |
| "fddt_use_overlap": true, |
| "fddt_use_silence": true, |
| "fddt_use_target": true, |
| "final_dropout": 0.0, |
| "forced_decoder_ids": null, |
| "init_std": 0.02, |
| "is_encoder_decoder": true, |
| "mask_feature_length": 10, |
| "mask_feature_min_masks": 0, |
| "mask_feature_prob": 0.0, |
| "mask_time_length": 10, |
| "mask_time_min_masks": 2, |
| "mask_time_prob": 0.05, |
| "max_source_positions": 1500, |
| "max_target_positions": 448, |
| "median_filter_width": 7, |
| "model_type": "DiCoW", |
| "mt_num_speakers": 1, |
| "n_soft_prompts": 16, |
| "non_target_fddt_value": 0.5, |
| "num_hidden_layers": 32, |
| "num_mel_bins": 128, |
| "pad_token_id": 50257, |
| "remove_timestamps_from_ctc": false, |
| "scale_embedding": false, |
| "scb_layers": -1, |
| "scb_method": null, |
| "sub_sample": true, |
| "torch_dtype": "float32", |
| "transformers_version": "4.42.0", |
| "use_cache": true, |
| "use_fddt": true, |
| "use_initial_fddt": true, |
| "use_weighted_layer_sum": false, |
| "vocab_size": 51866 |
| } |
|
|