Hugging Face's logo

Aly116
/
whisper-small-ur

Automatic Speech Recognition
Transformers
TensorBoard
Safetensors
Urdu
whisper
Generated from Trainer
Model card Files
xet
 Metrics Community
Aly116 commited on
Commit
fbcfda6
·
verified ·
1 Parent(s): 26758e7

Training in progress, step 1000

Browse files
Files changed (6) hide show
  1. config.json +61 -61
  2. model.safetensors +1 -1
  3. preprocessor_config.json +14 -14
  4. runs/Jan26_02-09-42_DESKTOP-5O2ELUM/events.out.tfevents.1737839455.DESKTOP-5O2ELUM.27556.0 +3 -0
  5. runs/Jan26_02-09-42_DESKTOP-5O2ELUM/events.out.tfevents.1737839749.DESKTOP-5O2ELUM.27556.1 +3 -0
  6. training_args.bin +1 -1
config.json CHANGED
@@ -1,61 +1,61 @@
1
- {
2
- "_name_or_path": "openai/whisper-small",
3
- "activation_dropout": 0.0,
4
- "activation_function": "gelu",
5
- "apply_spec_augment": false,
6
- "architectures": [
7
- "WhisperForConditionalGeneration"
8
- ],
9
- "attention_dropout": 0.0,
10
- "begin_suppress_tokens": null,
11
- "bos_token_id": 50257,
12
- "classifier_proj_size": 256,
13
- "d_model": 768,
14
- "decoder_attention_heads": 12,
15
- "decoder_ffn_dim": 3072,
16
- "decoder_layerdrop": 0.0,
17
- "decoder_layers": 12,
18
- "decoder_start_token_id": 50258,
19
- "dropout": 0.0,
20
- "encoder_attention_heads": 12,
21
- "encoder_ffn_dim": 3072,
22
- "encoder_layerdrop": 0.0,
23
- "encoder_layers": 12,
24
- "eos_token_id": 50257,
25
- "forced_decoder_ids": [
26
- [
27
- 1,
28
- 50259
29
- ],
30
- [
31
- 2,
32
- 50359
33
- ],
34
- [
35
- 3,
36
- 50363
37
- ]
38
- ],
39
- "init_std": 0.02,
40
- "is_encoder_decoder": true,
41
- "mask_feature_length": 10,
42
- "mask_feature_min_masks": 0,
43
- "mask_feature_prob": 0.0,
44
- "mask_time_length": 10,
45
- "mask_time_min_masks": 2,
46
- "mask_time_prob": 0.05,
47
- "max_length": null,
48
- "max_source_positions": 1500,
49
- "max_target_positions": 448,
50
- "median_filter_width": 7,
51
- "model_type": "whisper",
52
- "num_hidden_layers": 12,
53
- "num_mel_bins": 80,
54
- "pad_token_id": 50257,
55
- "scale_embedding": false,
56
- "torch_dtype": "float32",
57
- "transformers_version": "4.48.1",
58
- "use_cache": true,
59
- "use_weighted_layer_sum": false,
60
- "vocab_size": 51865
61
- }
 
1
+ {
2
+ "_name_or_path": "openai/whisper-small",
3
+ "activation_dropout": 0.0,
4
+ "activation_function": "gelu",
5
+ "apply_spec_augment": false,
6
+ "architectures": [
7
+ "WhisperForConditionalGeneration"
8
+ ],
9
+ "attention_dropout": 0.0,
10
+ "begin_suppress_tokens": null,
11
+ "bos_token_id": 50257,
12
+ "classifier_proj_size": 256,
13
+ "d_model": 768,
14
+ "decoder_attention_heads": 12,
15
+ "decoder_ffn_dim": 3072,
16
+ "decoder_layerdrop": 0.0,
17
+ "decoder_layers": 12,
18
+ "decoder_start_token_id": 50258,
19
+ "dropout": 0.0,
20
+ "encoder_attention_heads": 12,
21
+ "encoder_ffn_dim": 3072,
22
+ "encoder_layerdrop": 0.0,
23
+ "encoder_layers": 12,
24
+ "eos_token_id": 50257,
25
+ "forced_decoder_ids": [
26
+ [
27
+ 1,
28
+ 50259
29
+ ],
30
+ [
31
+ 2,
32
+ 50359
33
+ ],
34
+ [
35
+ 3,
36
+ 50363
37
+ ]
38
+ ],
39
+ "init_std": 0.02,
40
+ "is_encoder_decoder": true,
41
+ "mask_feature_length": 10,
42
+ "mask_feature_min_masks": 0,
43
+ "mask_feature_prob": 0.0,
44
+ "mask_time_length": 10,
45
+ "mask_time_min_masks": 2,
46
+ "mask_time_prob": 0.05,
47
+ "max_length": null,
48
+ "max_source_positions": 1500,
49
+ "max_target_positions": 448,
50
+ "median_filter_width": 7,
51
+ "model_type": "whisper",
52
+ "num_hidden_layers": 12,
53
+ "num_mel_bins": 80,
54
+ "pad_token_id": 50257,
55
+ "scale_embedding": false,
56
+ "torch_dtype": "float32",
57
+ "transformers_version": "4.48.1",
58
+ "use_cache": true,
59
+ "use_weighted_layer_sum": false,
60
+ "vocab_size": 51865
61
+ }
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:37bc11c7aa97a0e9f6f504381f3d1fdd35af6f3aad752db9bdbe77ae74936cd9
3
  size 966995080
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:40e5b7809feab1891de600918d8e1a6033243ebc31210b71ae5b51886f0972c6
3
  size 966995080
preprocessor_config.json CHANGED
@@ -1,14 +1,14 @@
1
- {
2
- "chunk_length": 30,
3
- "feature_extractor_type": "WhisperFeatureExtractor",
4
- "feature_size": 80,
5
- "hop_length": 160,
6
- "n_fft": 400,
7
- "n_samples": 480000,
8
- "nb_max_frames": 3000,
9
- "padding_side": "right",
10
- "padding_value": 0.0,
11
- "processor_class": "WhisperProcessor",
12
- "return_attention_mask": false,
13
- "sampling_rate": 16000
14
- }
 
1
+ {
2
+ "chunk_length": 30,
3
+ "feature_extractor_type": "WhisperFeatureExtractor",
4
+ "feature_size": 80,
5
+ "hop_length": 160,
6
+ "n_fft": 400,
7
+ "n_samples": 480000,
8
+ "nb_max_frames": 3000,
9
+ "padding_side": "right",
10
+ "padding_value": 0.0,
11
+ "processor_class": "WhisperProcessor",
12
+ "return_attention_mask": false,
13
+ "sampling_rate": 16000
14
+ }
runs/Jan26_02-09-42_DESKTOP-5O2ELUM/events.out.tfevents.1737839455.DESKTOP-5O2ELUM.27556.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e45c2c8d1c563198fe0ad2718daeb123c7c69e78670f1381ff95e62f3ed4affb
3
+ size 6823
runs/Jan26_02-09-42_DESKTOP-5O2ELUM/events.out.tfevents.1737839749.DESKTOP-5O2ELUM.27556.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8995fab368b44fc40feb3080956c992b3cb01de429d9977ab4a735f9f51ed430
3
+ size 15561
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:5e9c8c5d4c24edc4f699d9df00fc09441520f47b1a7c0a2b052785555b946159
3
  size 5496
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:20ddd63f30d06341bae7a27c0aa572f7425ad623c3348eec16136bc7773b4470
3
  size 5496