File size: 1,740 Bytes
4abc1d2
908a208
4abc1d2
 
f9d18d1
 
4abc1d2
 
 
 
 
 
 
 
 
 
908a208
 
f9d18d1
 
 
4abc1d2
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
fe8245b
 
4abc1d2
 
908a208
 
4abc1d2
 
f9d18d1
 
4abc1d2
 
f9d18d1
 
 
 
 
 
 
 
 
 
4abc1d2
 
 
 
 
 
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
---
license: apache-2.0
tags:
- generated_from_trainer
metrics:
- wer
model-index:
- name: wav2vec2-large-asr-th-2
  results: []
---

<!-- 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. -->

# wav2vec2-large-asr-th-2

This model is a fine-tuned version of [facebook/wav2vec2-large-xlsr-53](https://huggingface.co/facebook/wav2vec2-large-xlsr-53) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 0.3300
- Wer: 0.4924
- Cer: 0.1398

## 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: 0.0002
- train_batch_size: 16
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 3
- total_train_batch_size: 48
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 600
- training_steps: 4000
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch | Step | Validation Loss | Wer    | Cer    |
|:-------------:|:-----:|:----:|:---------------:|:------:|:------:|
| 1.2666        | 0.24  | 1000 | 0.7343          | 0.6934 | 0.2410 |
| 0.9305        | 0.47  | 2000 | 0.4504          | 0.6006 | 0.1764 |
| 0.8241        | 0.71  | 3000 | 0.3653          | 0.5278 | 0.1547 |
| 0.7788        | 0.95  | 4000 | 0.3300          | 0.4924 | 0.1398 |


### Framework versions

- Transformers 4.27.3
- Pytorch 1.13.1+cu116
- Datasets 2.10.1
- Tokenizers 0.13.2