|
{ |
|
"activation_dropout": 0.1, |
|
"adapter_kernel_size": 3, |
|
"adapter_stride": 2, |
|
"add_adapter": false, |
|
"apply_spec_augment": false, |
|
"apply_time_warp": false, |
|
"attention_dropout": 0.1, |
|
"bos_token_id": 1, |
|
"classifier_proj_size": 256, |
|
"codevector_dim": 256, |
|
"contrastive_logits_temperature": 0.1, |
|
"conv_bias": false, |
|
"conv_dim": [ |
|
384, |
|
384 |
|
], |
|
"conv_kernel": [ |
|
3, |
|
3 |
|
], |
|
"conv_stride": [ |
|
2, |
|
2 |
|
], |
|
"csgu_activation": "identity", |
|
"csgu_conv_dropout": 0.1, |
|
"csgu_kernel_size": 31, |
|
"csgu_use_linear_after_conv": false, |
|
"ctc_loss_reduction": "sum", |
|
"ctc_zero_infinity": true, |
|
"diversity_loss_weight": 0.1, |
|
"ebranchformer_conv_dropout": 0.1, |
|
"eos_token_id": 2, |
|
"fe_position_embeddings": true, |
|
"feat_extract_activation": "gelu", |
|
"feat_extract_norm": "group", |
|
"feat_proj_dropout": 0.0, |
|
"feat_quantizer_dropout": 0.0, |
|
"final_dropout": 0.1, |
|
"hidden_act": "gelu", |
|
"hidden_dropout": 0.1, |
|
"hidden_size": 384, |
|
"initializer_range": 0.02, |
|
"intermediate_size": 1536, |
|
"layer_norm_eps": 1e-05, |
|
"layerdrop": 0.1, |
|
"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": 1024, |
|
"merge_conv_kernel": 31, |
|
"model_type": "wav2vec2-ebranchformer", |
|
"num_adapter_layers": 3, |
|
"num_attention_heads": 4, |
|
"num_codevector_groups": 2, |
|
"num_codevectors_per_group": 320, |
|
"num_conv_pos_embedding_groups": 16, |
|
"num_conv_pos_embeddings": 128, |
|
"num_feat_extract_layers": 2, |
|
"num_hidden_layers": 12, |
|
"num_mel_bins": 80, |
|
"num_negatives": 100, |
|
"output_hidden_size": 384, |
|
"pad_token_id": 0, |
|
"position_embeddings_type": "relative", |
|
"proj_codevector_dim": 256, |
|
"rotary_embedding_base": 10000, |
|
"tdnn_dilation": [ |
|
1, |
|
2, |
|
3, |
|
1, |
|
1 |
|
], |
|
"tdnn_dim": [ |
|
512, |
|
512, |
|
512, |
|
512, |
|
1500 |
|
], |
|
"tdnn_kernel": [ |
|
5, |
|
3, |
|
3, |
|
1, |
|
1 |
|
], |
|
"time_warp_mode": "bicubic", |
|
"time_warp_window": 5, |
|
"transformers_version": "4.31.0.dev0", |
|
"use_fbanks": true, |
|
"use_macaron_ff": true, |
|
"use_weighted_layer_sum": false, |
|
"vocab_size": null, |
|
"xvector_output_dim": 512 |
|
} |
|
|