config.json
1.8 KB · 82 lines · json Raw
1 {
2 "activation_dropout": 0.0,
3 "adapter_act": "relu",
4 "adapter_kernel_size": 3,
5 "adapter_stride": 2,
6 "add_adapter": false,
7 "apply_spec_augment": false,
8 "architectures": [
9 "Wav2Vec2BertModel"
10 ],
11 "attention_dropout": 0.0,
12 "bos_token_id": 1,
13 "classifier_proj_size": 768,
14 "codevector_dim": 768,
15 "conformer_conv_dropout": 0.1,
16 "contrastive_logits_temperature": 0.1,
17 "conv_depthwise_kernel_size": 31,
18 "ctc_loss_reduction": "sum",
19 "ctc_zero_infinity": false,
20 "diversity_loss_weight": 0.1,
21 "eos_token_id": 2,
22 "feat_proj_dropout": 0.0,
23 "feat_quantizer_dropout": 0.0,
24 "feature_projection_input_dim": 160,
25 "final_dropout": 0.1,
26 "hidden_act": "swish",
27 "hidden_dropout": 0.0,
28 "hidden_size": 1024,
29 "initializer_range": 0.02,
30 "intermediate_size": 4096,
31 "layer_norm_eps": 1e-05,
32 "layerdrop": 0.1,
33 "left_max_position_embeddings": 64,
34 "mask_feature_length": 10,
35 "mask_feature_min_masks": 0,
36 "mask_feature_prob": 0.0,
37 "mask_time_length": 10,
38 "mask_time_min_masks": 2,
39 "mask_time_prob": 0.05,
40 "max_source_positions": 5000,
41 "model_type": "wav2vec2-bert",
42 "num_adapter_layers": 1,
43 "num_attention_heads": 16,
44 "num_codevector_groups": 2,
45 "num_codevectors_per_group": 320,
46 "num_hidden_layers": 24,
47 "num_negatives": 100,
48 "output_hidden_size": 1024,
49 "pad_token_id": 0,
50 "position_embeddings_type": "relative_key",
51 "proj_codevector_dim": 768,
52 "right_max_position_embeddings": 8,
53 "rotary_embedding_base": 10000,
54 "tdnn_dilation": [
55 1,
56 2,
57 3,
58 1,
59 1
60 ],
61 "tdnn_dim": [
62 512,
63 512,
64 512,
65 512,
66 1500
67 ],
68 "tdnn_kernel": [
69 5,
70 3,
71 3,
72 1,
73 1
74 ],
75 "torch_dtype": "float32",
76 "transformers_version": "4.37.0.dev0",
77 "use_intermediate_ffn_before_adapter": false,
78 "use_weighted_layer_sum": false,
79 "vocab_size": null,
80 "xvector_output_dim": 512
81 }
82