bond005 commited on
Commit
06c5754
1 Parent(s): 04ad973

Upload 7 files

Browse files
Files changed (3) hide show
  1. config.json +4 -35
  2. pytorch_model.bin +2 -2
  3. tokenizer_config.json +2 -2
config.json CHANGED
@@ -1,16 +1,12 @@
1
  {
2
- "_name_or_path": "wav2vec2-large-ru-golos-with-lm",
3
  "activation_dropout": 0.0,
4
- "adapter_kernel_size": 3,
5
- "adapter_stride": 2,
6
- "add_adapter": false,
7
  "apply_spec_augment": true,
8
  "architectures": [
9
  "Wav2Vec2ForCTC"
10
  ],
11
  "attention_dropout": 0.1,
12
  "bos_token_id": 1,
13
- "classifier_proj_size": 256,
14
  "codevector_dim": 768,
15
  "contrastive_logits_temperature": 0.1,
16
  "conv_bias": true,
@@ -52,6 +48,7 @@
52
  "feat_proj_dropout": 0.0,
53
  "feat_quantizer_dropout": 0.0,
54
  "final_dropout": 0.0,
 
55
  "hidden_act": "gelu",
56
  "hidden_dropout": 0.1,
57
  "hidden_size": 1024,
@@ -65,16 +62,13 @@
65
  "mask_channel_prob": 0.0,
66
  "mask_channel_selection": "static",
67
  "mask_feature_length": 10,
68
- "mask_feature_min_masks": 0,
69
  "mask_feature_prob": 0.0,
70
  "mask_time_length": 10,
71
- "mask_time_min_masks": 2,
72
  "mask_time_min_space": 1,
73
  "mask_time_other": 0.0,
74
  "mask_time_prob": 0.05,
75
  "mask_time_selection": "static",
76
  "model_type": "wav2vec2",
77
- "num_adapter_layers": 3,
78
  "num_attention_heads": 16,
79
  "num_codevector_groups": 2,
80
  "num_codevectors_per_group": 320,
@@ -83,33 +77,8 @@
83
  "num_feat_extract_layers": 7,
84
  "num_hidden_layers": 24,
85
  "num_negatives": 100,
86
- "output_hidden_size": 1024,
87
  "pad_token_id": 0,
88
  "proj_codevector_dim": 768,
89
- "tdnn_dilation": [
90
- 1,
91
- 2,
92
- 3,
93
- 1,
94
- 1
95
- ],
96
- "tdnn_dim": [
97
- 512,
98
- 512,
99
- 512,
100
- 512,
101
- 1500
102
- ],
103
- "tdnn_kernel": [
104
- 5,
105
- 3,
106
- 3,
107
- 1,
108
- 1
109
- ],
110
- "torch_dtype": "float32",
111
- "transformers_version": "4.22.0",
112
- "use_weighted_layer_sum": false,
113
- "vocab_size": 37,
114
- "xvector_output_dim": 512
115
  }
 
1
  {
2
+ "_name_or_path": "models/wav2vec2-large-ru-golos__",
3
  "activation_dropout": 0.0,
 
 
 
4
  "apply_spec_augment": true,
5
  "architectures": [
6
  "Wav2Vec2ForCTC"
7
  ],
8
  "attention_dropout": 0.1,
9
  "bos_token_id": 1,
 
10
  "codevector_dim": 768,
11
  "contrastive_logits_temperature": 0.1,
12
  "conv_bias": true,
 
48
  "feat_proj_dropout": 0.0,
49
  "feat_quantizer_dropout": 0.0,
50
  "final_dropout": 0.0,
51
+ "gradient_checkpointing": true,
52
  "hidden_act": "gelu",
53
  "hidden_dropout": 0.1,
54
  "hidden_size": 1024,
 
62
  "mask_channel_prob": 0.0,
63
  "mask_channel_selection": "static",
64
  "mask_feature_length": 10,
 
65
  "mask_feature_prob": 0.0,
66
  "mask_time_length": 10,
 
67
  "mask_time_min_space": 1,
68
  "mask_time_other": 0.0,
69
  "mask_time_prob": 0.05,
70
  "mask_time_selection": "static",
71
  "model_type": "wav2vec2",
 
72
  "num_attention_heads": 16,
73
  "num_codevector_groups": 2,
74
  "num_codevectors_per_group": 320,
 
77
  "num_feat_extract_layers": 7,
78
  "num_hidden_layers": 24,
79
  "num_negatives": 100,
 
80
  "pad_token_id": 0,
81
  "proj_codevector_dim": 768,
82
+ "transformers_version": "4.5.1",
83
+ "vocab_size": 37
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
84
  }
pytorch_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:8175e932bb5a549412f033265b8f6382a2df4ecb6ef5647a03dd78a9a2d777bd
3
- size 1262044529
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:19c82cbf135e3cfbfeecaf1886845de72ec2e38b3145d5d296981e2d1565c6dc
3
+ size 1262075377
tokenizer_config.json CHANGED
@@ -2,11 +2,11 @@
2
  "bos_token": "<s>",
3
  "do_lower_case": false,
4
  "eos_token": "</s>",
5
- "name_or_path": "/home/bond005/models/wav2vec2-large-ru-golos-with-lm",
6
  "pad_token": "<pad>",
7
  "processor_class": "Wav2Vec2ProcessorWithLM",
8
  "replace_word_delimiter_char": " ",
9
- "special_tokens_map_file": "/home/bond005/models/wav2vec2-large-ru-golos-with-lm/special_tokens_map.json",
10
  "tokenizer_class": "Wav2Vec2CTCTokenizer",
11
  "unk_token": "<unk>",
12
  "word_delimiter_token": "|"
 
2
  "bos_token": "<s>",
3
  "do_lower_case": false,
4
  "eos_token": "</s>",
5
+ "name_or_path": "/home/bond005/models/wav2vec2-large-ru-golos-with-lm2",
6
  "pad_token": "<pad>",
7
  "processor_class": "Wav2Vec2ProcessorWithLM",
8
  "replace_word_delimiter_char": " ",
9
+ "special_tokens_map_file": "/home/bond005/models/wav2vec2-large-ru-golos-with-lm2/special_tokens_map.json",
10
  "tokenizer_class": "Wav2Vec2CTCTokenizer",
11
  "unk_token": "<unk>",
12
  "word_delimiter_token": "|"