Hugging Face's logo

brever
/
wav2vec2-base-demo-colab

Automatic Speech Recognition
Transformers
PyTorch
TensorBoard
wav2vec2
Generated from Trainer
Inference Endpoints
Model card Files Metrics Community
brever commited on
Commit
77494af
•
1 Parent(s): 131712d

Training in progress, step 500

Browse files
Files changed (17) hide show
  1. .gitignore +1 -0
  2. config.json +88 -0
  3. preprocessor_config.json +9 -0
  4. pytorch_model.bin +3 -0
  5. runs/May22_00-35-44_7a616dd3ae5b/1653179860.2203445/events.out.tfevents.1653179860.7a616dd3ae5b.73.1 +3 -0
  6. runs/May22_00-35-44_7a616dd3ae5b/events.out.tfevents.1653179860.7a616dd3ae5b.73.0 +3 -0
  7. runs/May22_00-52-41_7a616dd3ae5b/1653180865.1151402/events.out.tfevents.1653180865.7a616dd3ae5b.73.3 +3 -0
  8. runs/May22_00-52-41_7a616dd3ae5b/events.out.tfevents.1653180865.7a616dd3ae5b.73.2 +3 -0
  9. runs/May22_01-32-37_7a616dd3ae5b/1653183251.7598846/events.out.tfevents.1653183251.7a616dd3ae5b.73.5 +3 -0
  10. runs/May22_01-32-37_7a616dd3ae5b/events.out.tfevents.1653183251.7a616dd3ae5b.73.4 +3 -0
  11. runs/May22_01-45-32_7a616dd3ae5b/1653184024.0544403/events.out.tfevents.1653184024.7a616dd3ae5b.73.7 +3 -0
  12. runs/May22_01-45-32_7a616dd3ae5b/1653184384.4057038/events.out.tfevents.1653184384.7a616dd3ae5b.73.9 +3 -0
  13. runs/May22_01-45-32_7a616dd3ae5b/events.out.tfevents.1653184024.7a616dd3ae5b.73.6 +3 -0
  14. runs/May22_01-45-32_7a616dd3ae5b/events.out.tfevents.1653184384.7a616dd3ae5b.73.8 +3 -0
  15. runs/May22_01-55-14_7a616dd3ae5b/1653184603.7438166/events.out.tfevents.1653184603.7a616dd3ae5b.73.11 +3 -0
  16. runs/May22_01-55-14_7a616dd3ae5b/events.out.tfevents.1653184603.7a616dd3ae5b.73.10 +3 -0
  17. training_args.bin +3 -0
.gitignore ADDED
@@ -0,0 +1 @@
 
 
1
+ checkpoint-*/
config.json ADDED
@@ -0,0 +1,88 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "facebook/wav2vec2-base",
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
+ "classifier_proj_size": 256,
11
+ "codevector_dim": 256,
12
+ "contrastive_logits_temperature": 0.1,
13
+ "conv_bias": false,
14
+ "conv_dim": [
15
+ 512,
16
+ 512,
17
+ 512,
18
+ 512,
19
+ 512,
20
+ 512,
21
+ 512
22
+ ],
23
+ "conv_kernel": [
24
+ 10,
25
+ 3,
26
+ 3,
27
+ 3,
28
+ 3,
29
+ 2,
30
+ 2
31
+ ],
32
+ "conv_stride": [
33
+ 5,
34
+ 2,
35
+ 2,
36
+ 2,
37
+ 2,
38
+ 2,
39
+ 2
40
+ ],
41
+ "ctc_loss_reduction": "mean",
42
+ "ctc_zero_infinity": false,
43
+ "diversity_loss_weight": 0.1,
44
+ "do_stable_layer_norm": false,
45
+ "eos_token_id": 2,
46
+ "feat_extract_activation": "gelu",
47
+ "feat_extract_norm": "group",
48
+ "feat_proj_dropout": 0.1,
49
+ "feat_quantizer_dropout": 0.0,
50
+ "final_dropout": 0.0,
51
+ "freeze_feat_extract_train": true,
52
+ "hidden_act": "gelu",
53
+ "hidden_dropout": 0.1,
54
+ "hidden_size": 768,
55
+ "initializer_range": 0.02,
56
+ "intermediate_size": 3072,
57
+ "layer_norm_eps": 1e-05,
58
+ "layerdrop": 0.0,
59
+ "mask_channel_length": 10,
60
+ "mask_channel_min_space": 1,
61
+ "mask_channel_other": 0.0,
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
+ "no_mask_channel_overlap": false,
73
+ "no_mask_time_overlap": false,
74
+ "num_attention_heads": 12,
75
+ "num_codevector_groups": 2,
76
+ "num_codevectors_per_group": 320,
77
+ "num_conv_pos_embedding_groups": 16,
78
+ "num_conv_pos_embeddings": 128,
79
+ "num_feat_extract_layers": 7,
80
+ "num_hidden_layers": 12,
81
+ "num_negatives": 100,
82
+ "pad_token_id": 29,
83
+ "proj_codevector_dim": 256,
84
+ "torch_dtype": "float32",
85
+ "transformers_version": "4.11.3",
86
+ "use_weighted_layer_sum": false,
87
+ "vocab_size": 32
88
+ }
preprocessor_config.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_normalize": true,
3
+ "feature_extractor_type": "Wav2Vec2FeatureExtractor",
4
+ "feature_size": 1,
5
+ "padding_side": "right",
6
+ "padding_value": 0.0,
7
+ "return_attention_mask": false,
8
+ "sampling_rate": 16000
9
+ }
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4ab861be87eebd0fa7d7d9a765fd58ea8d1c440eca78414484d6883c1d761db1
3
+ size 377670039
runs/May22_00-35-44_7a616dd3ae5b/1653179860.2203445/events.out.tfevents.1653179860.7a616dd3ae5b.73.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7eed70328cd57e0d90f8bab57525d559528f933b601d93feed1ec2803b2fe07b
3
+ size 4537
runs/May22_00-35-44_7a616dd3ae5b/events.out.tfevents.1653179860.7a616dd3ae5b.73.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6c8738cb20a3da4aa9e0cc4b4a27e100ce26d1347bff2befc2ebe75993984c73
3
+ size 4446
runs/May22_00-52-41_7a616dd3ae5b/1653180865.1151402/events.out.tfevents.1653180865.7a616dd3ae5b.73.3 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:26844fd54d61b8e5c3e849141a5f1bb7a2f92862e08fd706eeeec766481feb83
3
+ size 4537
runs/May22_00-52-41_7a616dd3ae5b/events.out.tfevents.1653180865.7a616dd3ae5b.73.2 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:659aa67799682399590df1a9c63c8eb72a10f90d76d776075e0d458b6d359733
3
+ size 4446
runs/May22_01-32-37_7a616dd3ae5b/1653183251.7598846/events.out.tfevents.1653183251.7a616dd3ae5b.73.5 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c8890d7d5f21a691413619c2e07fb9836b208c18d61d02fa17d8a57e913dc979
3
+ size 4537
runs/May22_01-32-37_7a616dd3ae5b/events.out.tfevents.1653183251.7a616dd3ae5b.73.4 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:46203d33c330cd2250339369945dab50f2fcbe037c125ac294b8a02226c0be2c
3
+ size 4446
runs/May22_01-45-32_7a616dd3ae5b/1653184024.0544403/events.out.tfevents.1653184024.7a616dd3ae5b.73.7 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9231360837dd6f506496d9ca98645d21d71aa2aad8cd90408bb39b264cf7eeef
3
+ size 4537
runs/May22_01-45-32_7a616dd3ae5b/1653184384.4057038/events.out.tfevents.1653184384.7a616dd3ae5b.73.9 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:98b786015af793c7c2df7ddd913d12e838f25eccc71385fde47ad9514d973e10
3
+ size 4537
runs/May22_01-45-32_7a616dd3ae5b/events.out.tfevents.1653184024.7a616dd3ae5b.73.6 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:863a1834b3422714d849d2bbb5c87c00373d367cca9354609de1262abb7c7a32
3
+ size 4446
runs/May22_01-45-32_7a616dd3ae5b/events.out.tfevents.1653184384.7a616dd3ae5b.73.8 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6f32d62f519037d7218d725a4bfac355783bdeab26913aab10ee3302a9386843
3
+ size 4446
runs/May22_01-55-14_7a616dd3ae5b/1653184603.7438166/events.out.tfevents.1653184603.7a616dd3ae5b.73.11 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3811d425359fb1eb26e73b95a4cb7f542e7dc9857797b11875c870ad95d9bccc
3
+ size 4537
runs/May22_01-55-14_7a616dd3ae5b/events.out.tfevents.1653184603.7a616dd3ae5b.73.10 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5882659dec95401691ec54f8617c76e072ecb605030142cffdfb6b7c4e1eecd3
3
+ size 4922
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b32ca16a71a97474892bcf70640bd933dfcda9d17fb9d92c5d88a9f2c7566820
3
+ size 2863