File size: 2,304 Bytes
f535954
d5183eb
 
 
 
 
 
 
 
 
 
04687ec
7483787
d5183eb
 
 
 
 
 
 
7483787
04687ec
7483787
f535954
 
d5183eb
 
f535954
d5183eb
f535954
04687ec
d5183eb
04687ec
 
f535954
d5183eb
f535954
d5183eb
f535954
d5183eb
f535954
d5183eb
f535954
d5183eb
f535954
d5183eb
f535954
d5183eb
f535954
d5183eb
f535954
d5183eb
04687ec
d5183eb
 
 
 
 
 
 
04687ec
d5183eb
 
f535954
d5183eb
f535954
d5183eb
 
04687ec
 
 
 
 
 
 
 
 
 
 
f535954
 
d5183eb
f535954
d5183eb
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
---
tags:
- generated_from_trainer
datasets:
- common_voice_17_0
metrics:
- wer
model-index:
- name: w2v2-bert-urdu
  results:
  - task:
      type: automatic-speech-recognition
      name: Automatic Speech Recognition
    dataset:
      name: common_voice_17_0
      type: common_voice_17_0
      config: ur
      split: test[:100]
      args: ur
    metrics:
    - type: wer
      value: 0.3300546448087432
      name: Wer
---

<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->

# w2v2-bert-urdu

This model was trained from scratch on the common_voice_17_0 dataset.
It achieves the following results on the evaluation set:
- Loss: 0.4246
- Wer: 0.3301

## Model description

More information needed

## Intended uses & limitations

More information needed

## Training and evaluation data

More information needed

## Training procedure

### Training hyperparameters

The following hyperparameters were used during training:
- learning_rate: 5e-06
- train_batch_size: 16
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 32
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 100
- num_epochs: 2
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch  | Step | Validation Loss | Wer    |
|:-------------:|:------:|:----:|:---------------:|:------:|
| 0.8145        | 0.1695 | 50   | 0.4620          | 0.3421 |
| 0.4364        | 0.3390 | 100  | 0.3969          | 0.2874 |
| 0.418         | 0.5085 | 150  | 0.3697          | 0.2820 |
| 0.402         | 0.6780 | 200  | 0.3627          | 0.2842 |
| 0.3698        | 0.8475 | 250  | 0.3314          | 0.2710 |
| 0.3779        | 1.0169 | 300  | 0.3292          | 0.2852 |
| 0.3167        | 1.1864 | 350  | 0.3230          | 0.2820 |
| 0.3578        | 1.3559 | 400  | 0.3825          | 0.2940 |
| 0.4189        | 1.5254 | 450  | 0.4225          | 0.3104 |
| 0.4803        | 1.6949 | 500  | 0.4248          | 0.3311 |
| 0.4612        | 1.8644 | 550  | 0.4246          | 0.3301 |


### Framework versions

- Transformers 4.40.2
- Pytorch 2.2.1+cu121
- Datasets 2.19.1
- Tokenizers 0.19.1