orf_gu_wav2vec2_v2 / config.json
Jatin-WIAI's picture
model trained on 11-01-2024 https://docs.google.com/spreadsheets/d/1bAJNajr38MmwOyLq2ntgdOQM2mZZMiuk8hV0Qu9CqmQ/edit#gid=0&range=B43
7bdacef
{
"_name_or_path": "/home/users/jatin/models/asr/gu/pseudo_v2_ft_overlapping_v1_lora/checkpoint-9600",
"activation_dropout": 0.1,
"adapter_kernel_size": 3,
"adapter_name": "bottleneck",
"adapter_stride": 2,
"add_adapter": false,
"apply_spec_augment": true,
"architectures": [
"Wav2Vec2ForCTC"
],
"attention_dropout": 0.1,
"bos_token_id": 1,
"classifier_proj_size": 256,
"codevector_dim": 256,
"contrastive_logits_temperature": 0.1,
"conv_bias": true,
"conv_dim": [
512,
512,
512,
512,
512,
512,
512
],
"conv_kernel": [
10,
3,
3,
3,
3,
2,
2
],
"conv_stride": [
5,
2,
2,
2,
2,
2,
2
],
"ctc_loss_reduction": "mean",
"ctc_zero_infinity": false,
"diversity_loss_weight": 0.1,
"do_stable_layer_norm": true,
"eos_token_id": 2,
"feat_enc_adapter": false,
"feat_extract_activation": "gelu",
"feat_extract_dropout": 0.0,
"feat_extract_norm": "layer",
"feat_proj_dropout": 0.0,
"feat_quantizer_dropout": 0.0,
"final_dropout": 0.1,
"hidden_act": "gelu",
"hidden_dropout": 0.1,
"hidden_dropout_prob": 0.1,
"hidden_size": 1024,
"initializer_range": 0.02,
"intermediate_size": 4096,
"layer_norm_eps": 1e-05,
"layerdrop": 0.1,
"lora_adapter": true,
"lora_r": 8,
"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,
"mh_adapter": false,
"model_type": "wav2vec2",
"num_adapter_layers": 3,
"num_attention_heads": 16,
"num_codevector_groups": 2,
"num_codevectors_per_group": 320,
"num_conv_pos_embedding_groups": 16,
"num_conv_pos_embeddings": 128,
"num_feat_extract_layers": 7,
"num_hidden_layers": 24,
"num_negatives": 100,
"output_adapter": false,
"output_hidden_size": 1024,
"pad_token_id": 0,
"prefix_tuning": false,
"proj_codevector_dim": 256,
"tdnn_dilation": [
1,
2,
3,
1,
1
],
"tdnn_dim": [
512,
512,
512,
512,
1500
],
"tdnn_kernel": [
5,
3,
3,
1,
1
],
"torch_dtype": "float32",
"transformers_version": "4.21.3",
"use_weighted_layer_sum": false,
"vocab_size": 70,
"xvector_output_dim": 512
}