File size: 2,282 Bytes
21880ea
236e09b
 
21880ea
 
236e09b
 
21880ea
 
 
 
 
 
 
 
 
 
 
 
236e09b
21880ea
 
 
236e09b
21880ea
 
 
 
 
 
 
 
 
 
 
236e09b
21880ea
236e09b
21880ea
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
---
language:
- vi
base_model: ylacombe/w2v-bert-2.0
tags:
- automatic-speech-recognition
- mozilla-foundation/common_voice_16_0
- generated_from_trainer
datasets:
- common_voice_16_0
metrics:
- wer
model-index:
- name: wav2vec2-common_voice-vi-demo
  results:
  - task:
      name: Automatic Speech Recognition
      type: automatic-speech-recognition
    dataset:
      name: MOZILLA-FOUNDATION/COMMON_VOICE_16_0 - VI
      type: common_voice_16_0
      config: vi
      split: test
      args: 'Config: vi, Training split: train+validation, Eval split: test'
    metrics:
    - name: Wer
      type: wer
      value: 1.0
---

<!-- 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-common_voice-vi-demo

This model is a fine-tuned version of [ylacombe/w2v-bert-2.0](https://huggingface.co/ylacombe/w2v-bert-2.0) on the MOZILLA-FOUNDATION/COMMON_VOICE_16_0 - VI dataset.
It achieves the following results on the evaluation set:
- Loss: 3.3958
- Wer: 1.0

## 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: 0.002
- 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: 500
- num_epochs: 15.0
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch | Step | Validation Loss | Wer |
|:-------------:|:-----:|:----:|:---------------:|:---:|
| No log        | 2.26  | 200  | 3.5924          | 1.0 |
| No log        | 4.52  | 400  | 3.4946          | 1.0 |
| 5.7152        | 6.78  | 600  | 3.4630          | 1.0 |
| 5.7152        | 9.04  | 800  | 3.4525          | 1.0 |
| 3.5048        | 11.3  | 1000 | 3.4329          | 1.0 |
| 3.5048        | 13.56 | 1200 | 3.4074          | 1.0 |


### Framework versions

- Transformers 4.37.0.dev0
- Pytorch 2.1.0+cu121
- Datasets 2.14.5
- Tokenizers 0.15.0