config.json
2.4 KB · 133 lines · json Raw
1 {
2 "_name_or_path": "onecxi/open-vakgyata",
3 "activation_dropout": 0.05,
4 "adapter_attn_dim": 16,
5 "adapter_kernel_size": 3,
6 "adapter_stride": 2,
7 "add_adapter": false,
8 "apply_spec_augment": true,
9 "architectures": [
10 "Wav2Vec2ForSequenceClassification"
11 ],
12 "attention_dropout": 0.05,
13 "bos_token_id": 1,
14 "classifier_proj_size": 1024,
15 "codevector_dim": 1024,
16 "contrastive_logits_temperature": 0.1,
17 "conv_bias": true,
18 "conv_dim": [
19 512,
20 512,
21 512,
22 512,
23 512,
24 512,
25 512
26 ],
27 "conv_kernel": [
28 10,
29 3,
30 3,
31 3,
32 3,
33 2,
34 2
35 ],
36 "conv_stride": [
37 5,
38 2,
39 2,
40 2,
41 2,
42 2,
43 2
44 ],
45 "ctc_loss_reduction": "mean",
46 "ctc_zero_infinity": false,
47 "diversity_loss_weight": 0.1,
48 "do_stable_layer_norm": true,
49 "eos_token_id": 2,
50 "feat_extract_activation": "gelu",
51 "feat_extract_dropout": 0.0,
52 "feat_extract_norm": "layer",
53 "feat_proj_dropout": 0.05,
54 "feat_quantizer_dropout": 0.0,
55 "final_dropout": 0.05,
56 "hidden_act": "gelu",
57 "hidden_dropout": 0.05,
58 "hidden_size": 1280,
59 "id2label": {
60 "0": "en-IN",
61 "1": "hi-IN",
62 "2": "or-IN",
63 "3": "bn-IN",
64 "4": "ta-IN",
65 "5": "te-IN",
66 "6": "kn-IN",
67 "7": "ml-IN",
68 "8": "mr-IN",
69 "9": "gu-IN"
70 },
71 "initializer_range": 0.02,
72 "intermediate_size": 5120,
73 "label2id": {
74 "bn-IN": 3,
75 "en-IN": 0,
76 "gu-IN": 9,
77 "hi-IN": 1,
78 "kn-IN": 6,
79 "ml-IN": 7,
80 "mr-IN": 8,
81 "or-IN": 2,
82 "ta-IN": 4,
83 "te-IN": 5
84 },
85 "layer_norm_eps": 1e-05,
86 "layerdrop": 0.05,
87 "mask_feature_length": 10,
88 "mask_feature_min_masks": 0,
89 "mask_feature_prob": 0.0,
90 "mask_time_length": 10,
91 "mask_time_min_masks": 2,
92 "mask_time_prob": 0.05,
93 "model_type": "wav2vec2",
94 "num_adapter_layers": 3,
95 "num_attention_heads": 16,
96 "num_codevector_groups": 2,
97 "num_codevectors_per_group": 320,
98 "num_conv_pos_embedding_groups": 16,
99 "num_conv_pos_embeddings": 128,
100 "num_feat_extract_layers": 7,
101 "num_hidden_layers": 2,
102 "num_negatives": 100,
103 "output_hidden_size": 1280,
104 "pad_token_id": 0,
105 "proj_codevector_dim": 1024,
106 "tdnn_dilation": [
107 1,
108 2,
109 3,
110 1,
111 1
112 ],
113 "tdnn_dim": [
114 512,
115 512,
116 512,
117 512,
118 1500
119 ],
120 "tdnn_kernel": [
121 5,
122 3,
123 3,
124 1,
125 1
126 ],
127 "torch_dtype": "float32",
128 "transformers_version": "4.42.0",
129 "use_weighted_layer_sum": false,
130 "vocab_size": 154,
131 "xvector_output_dim": 512
132 }
133