syedmuhammad commited on
Commit
a81f990
1 Parent(s): e430f90

second_release

Browse files
README.md CHANGED
@@ -5,9 +5,24 @@ tags:
5
  - generated_from_trainer
6
  datasets:
7
  - common_voice_13_0
 
 
8
  model-index:
9
  - name: Wav2Vec2-Urdu-300M
10
- results: []
 
 
 
 
 
 
 
 
 
 
 
 
 
11
  ---
12
 
13
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -17,13 +32,8 @@ should probably proofread and complete it, then remove this comment. -->
17
 
18
  This model is a fine-tuned version of [facebook/wav2vec2-large-xlsr-53](https://huggingface.co/facebook/wav2vec2-large-xlsr-53) on the common_voice_13_0 dataset.
19
  It achieves the following results on the evaluation set:
20
- - eval_loss: inf
21
- - eval_wer: 0.3928
22
- - eval_runtime: 234.5067
23
- - eval_samples_per_second: 14.089
24
- - eval_steps_per_second: 1.761
25
- - epoch: 13.56
26
- - step: 8400
27
 
28
  ## Model description
29
 
@@ -43,17 +53,28 @@ More information needed
43
 
44
  The following hyperparameters were used during training:
45
  - learning_rate: 0.0003
46
- - train_batch_size: 6
47
  - eval_batch_size: 8
48
  - seed: 42
49
  - gradient_accumulation_steps: 2
50
- - total_train_batch_size: 12
51
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
52
  - lr_scheduler_type: linear
53
  - lr_scheduler_warmup_steps: 500
54
- - num_epochs: 18
55
  - mixed_precision_training: Native AMP
56
 
 
 
 
 
 
 
 
 
 
 
 
57
  ### Framework versions
58
 
59
  - Transformers 4.35.0
 
5
  - generated_from_trainer
6
  datasets:
7
  - common_voice_13_0
8
+ metrics:
9
+ - wer
10
  model-index:
11
  - name: Wav2Vec2-Urdu-300M
12
+ results:
13
+ - task:
14
+ name: Automatic Speech Recognition
15
+ type: automatic-speech-recognition
16
+ dataset:
17
+ name: common_voice_13_0
18
+ type: common_voice_13_0
19
+ config: ur
20
+ split: test
21
+ args: ur
22
+ metrics:
23
+ - name: Wer
24
+ type: wer
25
+ value: 0.382515146807519
26
  ---
27
 
28
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
 
32
 
33
  This model is a fine-tuned version of [facebook/wav2vec2-large-xlsr-53](https://huggingface.co/facebook/wav2vec2-large-xlsr-53) on the common_voice_13_0 dataset.
34
  It achieves the following results on the evaluation set:
35
+ - Loss: inf
36
+ - Wer: 0.3825
 
 
 
 
 
37
 
38
  ## Model description
39
 
 
53
 
54
  The following hyperparameters were used during training:
55
  - learning_rate: 0.0003
56
+ - train_batch_size: 8
57
  - eval_batch_size: 8
58
  - seed: 42
59
  - gradient_accumulation_steps: 2
60
+ - total_train_batch_size: 16
61
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
62
  - lr_scheduler_type: linear
63
  - lr_scheduler_warmup_steps: 500
64
+ - num_epochs: 14
65
  - mixed_precision_training: Native AMP
66
 
67
+ ### Training results
68
+
69
+ | Training Loss | Epoch | Step | Validation Loss | Wer |
70
+ |:-------------:|:-----:|:----:|:---------------:|:------:|
71
+ | 0.9658 | 2.58 | 1200 | inf | 0.5573 |
72
+ | 0.5669 | 5.17 | 2400 | inf | 0.4473 |
73
+ | 0.4117 | 7.75 | 3600 | inf | 0.4121 |
74
+ | 0.3197 | 10.33 | 4800 | inf | 0.4039 |
75
+ | 0.2667 | 12.92 | 6000 | inf | 0.3825 |
76
+
77
+
78
  ### Framework versions
79
 
80
  - Transformers 4.35.0
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:570d99f4ed2868b01becc38a8c316fc27d0db0818ec0c2e620340d391ddebd3c
3
  size 1261996080
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:deddac38042aa9d5e1cc81208a8ee860edc9ce71c185a8267a7c46246164e1f2
3
  size 1261996080
runs/Nov10_14-35-52_9ef673b706cd/events.out.tfevents.1699627195.9ef673b706cd.1359.0 CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:0aa4d2cf17b9bd201ee2ace52dc265abb9d4da2315ae9f247ac910462024829d
3
- size 9991
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5bd979b9b76b2e9802435d7e0a42bb9e14967faff27362d59d6f4f3bf89acedc
3
+ size 10345