Hugging Face's logo

Paywinful
/
hugging_dir

Safetensors
wav2vec2-bert
Model card Files
xet
Community
Paywinful commited on
Commit
87c6414
·
verified ·
1 Parent(s): 740e970

Training in progress, step 150

Browse files
This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. added_tokens.json +4 -0
  2. config.json +81 -0
  3. model.safetensors +3 -0
  4. preprocessor_config.json +11 -0
  5. special_tokens_map.json +6 -0
  6. test/data-00000-of-00001.arrow +3 -0
  7. test/dataset_info.json +15 -0
  8. test/state.json +13 -0
  9. tokenizer_config.json +49 -0
  10. train/data-00000-of-00059.arrow +3 -0
  11. train/data-00001-of-00059.arrow +3 -0
  12. train/data-00002-of-00059.arrow +3 -0
  13. train/data-00003-of-00059.arrow +3 -0
  14. train/data-00004-of-00059.arrow +3 -0
  15. train/data-00005-of-00059.arrow +3 -0
  16. train/data-00006-of-00059.arrow +3 -0
  17. train/data-00007-of-00059.arrow +3 -0
  18. train/data-00008-of-00059.arrow +3 -0
  19. train/data-00009-of-00059.arrow +3 -0
  20. train/data-00010-of-00059.arrow +3 -0
  21. train/data-00011-of-00059.arrow +3 -0
  22. train/data-00012-of-00059.arrow +3 -0
  23. train/data-00013-of-00059.arrow +3 -0
  24. train/data-00014-of-00059.arrow +3 -0
  25. train/data-00015-of-00059.arrow +3 -0
  26. train/data-00016-of-00059.arrow +3 -0
  27. train/data-00017-of-00059.arrow +3 -0
  28. train/data-00018-of-00059.arrow +3 -0
  29. train/data-00019-of-00059.arrow +3 -0
  30. train/data-00020-of-00059.arrow +3 -0
  31. train/data-00021-of-00059.arrow +3 -0
  32. train/data-00022-of-00059.arrow +3 -0
  33. train/data-00023-of-00059.arrow +3 -0
  34. train/data-00024-of-00059.arrow +3 -0
  35. train/data-00025-of-00059.arrow +3 -0
  36. train/data-00026-of-00059.arrow +3 -0
  37. train/data-00027-of-00059.arrow +3 -0
  38. train/data-00028-of-00059.arrow +3 -0
  39. train/data-00029-of-00059.arrow +3 -0
  40. train/data-00030-of-00059.arrow +3 -0
  41. train/data-00031-of-00059.arrow +3 -0
  42. train/data-00032-of-00059.arrow +3 -0
  43. train/data-00033-of-00059.arrow +3 -0
  44. train/data-00034-of-00059.arrow +3 -0
  45. train/data-00035-of-00059.arrow +3 -0
  46. train/data-00036-of-00059.arrow +3 -0
  47. train/data-00037-of-00059.arrow +3 -0
  48. train/data-00038-of-00059.arrow +3 -0
  49. train/data-00039-of-00059.arrow +3 -0
  50. train/data-00040-of-00059.arrow +3 -0
added_tokens.json ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ {
2
+ "</s>": 67,
3
+ "<s>": 66
4
+ }
config.json ADDED
@@ -0,0 +1,81 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "activation_dropout": 0.0,
3
+ "adapter_act": "relu",
4
+ "adapter_kernel_size": 3,
5
+ "adapter_stride": 2,
6
+ "add_adapter": true,
7
+ "apply_spec_augment": false,
8
+ "architectures": [
9
+ "Wav2Vec2BertForCTC"
10
+ ],
11
+ "attention_dropout": 0.1,
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": "mean",
19
+ "ctc_zero_infinity": false,
20
+ "diversity_loss_weight": 0.1,
21
+ "eos_token_id": 2,
22
+ "feat_proj_dropout": 0.1,
23
+ "feat_quantizer_dropout": 0.0,
24
+ "feature_projection_input_dim": 160,
25
+ "final_dropout": 0.1,
26
+ "hidden_act": "swish",
27
+ "hidden_dropout": 0.1,
28
+ "hidden_size": 1024,
29
+ "initializer_range": 0.02,
30
+ "intermediate_size": 4096,
31
+ "layer_norm_eps": 1e-05,
32
+ "layerdrop": 0.0,
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": 65,
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": "4.54.0.dev0",
77
+ "use_intermediate_ffn_before_adapter": false,
78
+ "use_weighted_layer_sum": false,
79
+ "vocab_size": 68,
80
+ "xvector_output_dim": 512
81
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a546e2452229e8a8520f15de92500004f80193c7fe8a7d535968cd39d20c3e2e
3
+ size 2423097560
preprocessor_config.json ADDED
@@ -0,0 +1,11 @@
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "feature_extractor_type": "SeamlessM4TFeatureExtractor",
3
+ "feature_size": 80,
4
+ "num_mel_bins": 80,
5
+ "padding_side": "right",
6
+ "padding_value": 1,
7
+ "processor_class": "Wav2Vec2BertProcessor",
8
+ "return_attention_mask": true,
9
+ "sampling_rate": 16000,
10
+ "stride": 2
11
+ }
special_tokens_map.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<s>",
3
+ "eos_token": "</s>",
4
+ "pad_token": "[PAD]",
5
+ "unk_token": "[UNK]"
6
+ }
test/data-00000-of-00001.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:21e50f8beb3ba74c924938ef4b875560557759704abe4d4b3325e84d9683121a
3
+ size 465901864
test/dataset_info.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "citation": "",
3
+ "description": "",
4
+ "features": {
5
+ "AUDIO_PATH": {
6
+ "_type": "Audio"
7
+ },
8
+ "Transcription": {
9
+ "dtype": "string",
10
+ "_type": "Value"
11
+ }
12
+ },
13
+ "homepage": "",
14
+ "license": ""
15
+ }
test/state.json ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_data_files": [
3
+ {
4
+ "filename": "data-00000-of-00001.arrow"
5
+ }
6
+ ],
7
+ "_fingerprint": "5e34afa7c5ea9c0e",
8
+ "_format_columns": null,
9
+ "_format_kwargs": {},
10
+ "_format_type": null,
11
+ "_output_all_columns": false,
12
+ "_split": null
13
+ }
tokenizer_config.json ADDED
@@ -0,0 +1,49 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "64": {
4
+ "content": "[UNK]",
5
+ "lstrip": true,
6
+ "normalized": false,
7
+ "rstrip": true,
8
+ "single_word": false,
9
+ "special": false
10
+ },
11
+ "65": {
12
+ "content": "[PAD]",
13
+ "lstrip": true,
14
+ "normalized": false,
15
+ "rstrip": true,
16
+ "single_word": false,
17
+ "special": false
18
+ },
19
+ "66": {
20
+ "content": "<s>",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "67": {
28
+ "content": "</s>",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ }
35
+ },
36
+ "bos_token": "<s>",
37
+ "clean_up_tokenization_spaces": false,
38
+ "do_lower_case": false,
39
+ "eos_token": "</s>",
40
+ "extra_special_tokens": {},
41
+ "model_max_length": 1000000000000000019884624838656,
42
+ "pad_token": "[PAD]",
43
+ "processor_class": "Wav2Vec2BertProcessor",
44
+ "replace_word_delimiter_char": " ",
45
+ "target_lang": null,
46
+ "tokenizer_class": "Wav2Vec2CTCTokenizer",
47
+ "unk_token": "[UNK]",
48
+ "word_delimiter_token": "|"
49
+ }
train/data-00000-of-00059.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5301c32fea0ad0d512ee9e3fc8b32d28f7d9ad2f39db32aa7bd19f3e8ff27a61
3
+ size 499476440
train/data-00001-of-00059.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:09aa9f7814bf8e8eb8e7a9b40642b07f91863c1501ac53cd7f6135d20ad57b81
3
+ size 496464632
train/data-00002-of-00059.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:414a83c6050af866ca5ef6b248985c9ebb234d0faa1617551104157797389005
3
+ size 495705040
train/data-00003-of-00059.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3413e8d190799d4329100710d5af03dd33bab4b4e7d73d1973cb585ba6839b59
3
+ size 498811536
train/data-00004-of-00059.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d592cad0f92fdc43f4aa258d37bc33be902d5b410ed79579aa8cca222cfda5fc
3
+ size 494294328
train/data-00005-of-00059.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6da3990eac54f0ad866f725742b952185ce39affdb71929c06ef0a9253684f3c
3
+ size 502141568
train/data-00006-of-00059.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9c2530fa74429d2f447a3b4ba8c4f46a9fadaaaeee9c15b5e1cfb52f544171c6
3
+ size 497870872
train/data-00007-of-00059.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3849ec812663b9ebb60d9128daeabd1bb3b5b90e3a958ab49dce73cbf22a39e8
3
+ size 499110264
train/data-00008-of-00059.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1a109ea2f4912f31c5c1eb471098e571aab72c3b1f5175d24cc141bbaf09bced
3
+ size 502462968
train/data-00009-of-00059.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:28495c8eb8a6a1f9ef5874a6099b4ae54c8e264208aefd56b454a655259a37ef
3
+ size 504714080
train/data-00010-of-00059.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6d47ce92091e6dd5de9f74b7dd6294cb70a96bc2061b25ab28372a909d6c506e
3
+ size 494345520
train/data-00011-of-00059.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c3dd488eb143a1998800d6706123f66a31a00fe4cf479c1dd79b485baf245894
3
+ size 495622816
train/data-00012-of-00059.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e058f7d8553a0894917cc4bce8627ce9c7fec576b54fdd03864459f7cf1a38f9
3
+ size 500819816
train/data-00013-of-00059.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ac4ced09e1ff863a29aefb1122cd38793f431c31d4eac6b107559eb3eb115199
3
+ size 501582080
train/data-00014-of-00059.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0ae3963c41fd183bfc21eacea4c728fae398dda5139e4ba0e51b742c3f4be9f2
3
+ size 501837888
train/data-00015-of-00059.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:68f0fac19dee05f65c566a7e07a90e4a0557a3337baa0bd5189d8f1da0defb45
3
+ size 492637736
train/data-00016-of-00059.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:10b18484ca230d6521b6f6ca720f38ed65c4ef7799c11f843f6de9afc30c85ce
3
+ size 498657496
train/data-00017-of-00059.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:28d5a5d7ea4f42268a815049315fe5f0113168dee3678002f47c654b2fd5724c
3
+ size 495965032
train/data-00018-of-00059.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1f0a34f9061bc9f87d6edd3cf71edb461abcd2da4fc605d98b4db32d86e0e37f
3
+ size 499673616
train/data-00019-of-00059.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1c596801a618f9760730ff1d076eef99a61c94207f9a71f635a1b58d25ed68b1
3
+ size 498442192
train/data-00020-of-00059.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2a61d581621dc5469f6b29c938115cc87112c78712bd96c481cef3f165625339
3
+ size 500413040
train/data-00021-of-00059.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d527a0cf45e27800d2780208edb4cefd21c8fdaf2e9abc1881944382e7e8081e
3
+ size 500106752
train/data-00022-of-00059.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:19db9dc7e7259e06c79c18a56918c4cdfe8b57ad601009859294ff50cded54d0
3
+ size 495207072
train/data-00023-of-00059.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b1048abbacee022f4007be0c19151a3087230642672e4a52b800b094e99612af
3
+ size 505953520
train/data-00024-of-00059.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:63b3b5c1726c1d8037bab544494be832285788cbc19ca26c9175f5d9112e1c4a
3
+ size 501308352
train/data-00025-of-00059.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:08d3c90918e6e1fe327a932ca20cdc884d31ecf9f5ec724383888dce4abcb132
3
+ size 497351976
train/data-00026-of-00059.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:616b504fc7114fb1e6f811500e467c83b7dd808c144acad7a43280cf1c89f5df
3
+ size 499477200
train/data-00027-of-00059.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:236fe7dd10e84bde5e3424c6832dae90e5bd1b2ec223bdb611f8e0d5e74a78ab
3
+ size 499893624
train/data-00028-of-00059.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:384fa86a5831d8908f25ef06bcdb63fd2ed532d961e06ae086b88ae52f58182e
3
+ size 496302176
train/data-00029-of-00059.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7dd041857551fc708ceadf7ee21540d4bf5ad7a99f50fcf959dea5bdead27388
3
+ size 495026976
train/data-00030-of-00059.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e0aa1ece5cea1f6d57e20aeb11c08ff2ec315787c6ac04e3c90fea3c4a4864f4
3
+ size 498280440
train/data-00031-of-00059.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e335fe6b8fd08841d476f553756e1a9c8063d433a1f8afa79e082a0c29f9f668
3
+ size 498250456
train/data-00032-of-00059.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9eeb761313ae51b4eb14761786f10a74014ff6d532a0c4071b4aa70ddebc0326
3
+ size 498291920
train/data-00033-of-00059.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dbd5bbabdff360b77504f74528e4041f15a65bc38d545cf83cf9580dbe56e481
3
+ size 501878328
train/data-00034-of-00059.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6a162d73ba36dfd31cb57fb2b9f718a60bf47c060e5559ea11e2521092d7ffad
3
+ size 497207600
train/data-00035-of-00059.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:844906124ac43108015a011d09054c889064616898c52c09b05823e3f0d36c95
3
+ size 497709968
train/data-00036-of-00059.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:74533d18ff784cecb24760266dc39c8abd5070e284d425122127daf5b1e7f9f7
3
+ size 503447320
train/data-00037-of-00059.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:efbf86dfa5aff8ce93bc8a9296b56059c19bc76f182494a90d65c49568404ef3
3
+ size 492689072
train/data-00038-of-00059.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:41b447795b2fa5e961c1e23039533fa671470d50a0d5aeedaa3523fc487b7d56
3
+ size 497066648
train/data-00039-of-00059.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0e20c9d75433fffe6f5c6233f34af406c368a8efd91b44de612c6c953f999f88
3
+ size 499872648
train/data-00040-of-00059.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0a37230b2635e5c29ab253e0b77be11fd092ac6a3833435f33fb1b9fafa359cf
3
+ size 494551952