File size: 3,018 Bytes
63d1e12
9ef2a89
 
 
4e5edfa
 
9ef2a89
 
 
4e5edfa
 
9ef2a89
 
 
 
27fcd84
4e5edfa
9ef2a89
 
 
 
7e61fbd
9ef2a89
 
4e5edfa
27fcd84
4e5edfa
 
27fcd84
4e5edfa
63d1e12
 
9ef2a89
 
63d1e12
9ef2a89
63d1e12
9ef2a89
 
27fcd84
 
 
 
63d1e12
9ef2a89
63d1e12
9ef2a89
63d1e12
9ef2a89
63d1e12
9ef2a89
63d1e12
9ef2a89
63d1e12
9ef2a89
63d1e12
9ef2a89
63d1e12
9ef2a89
63d1e12
9ef2a89
 
 
 
 
 
 
 
 
 
 
 
63d1e12
9ef2a89
63d1e12
27fcd84
 
 
 
 
 
 
 
 
 
 
 
 
 
 
63d1e12
 
9ef2a89
63d1e12
9ef2a89
 
 
 
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
89
90
91
92
93
94
95
96
97
98
99
---
base_model: facebook/mms-1b-all
datasets:
- common_voice_17_0
library_name: transformers
license: cc-by-nc-4.0
metrics:
- wer
- bleu
tags:
- generated_from_trainer
model-index:
- name: wav2vec2-mms-1b-CV17.0-training_set_variations
  results:
  - task:
      type: automatic-speech-recognition
      name: Automatic Speech Recognition
    dataset:
      name: common_voice_17_0
      type: common_voice_17_0
      config: ta
      split: validation
      args: ta
    metrics:
    - type: wer
      value: 0.3597180870859695
      name: Wer
    - type: bleu
      value: 0.4226157099926465
      name: Bleu
---

<!-- 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-mms-1b-CV17.0-training_set_variations

This model is a fine-tuned version of [facebook/mms-1b-all](https://huggingface.co/facebook/mms-1b-all) on the common_voice_17_0 dataset.
It achieves the following results on the evaluation set:
- Loss: 0.2047
- Wer: 0.3597
- Cer: 0.0579
- Bleu: 0.4226

## 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: 0.001
- 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_ratio: 0.15
- training_steps: 2000
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch  | Step | Validation Loss | Wer    | Cer    | Bleu   |
|:-------------:|:------:|:----:|:---------------:|:------:|:------:|:------:|
| 6.3615        | 0.3906 | 100  | 0.2954          | 0.4162 | 0.0682 | 0.3508 |
| 0.2115        | 0.7812 | 200  | 0.2266          | 0.3822 | 0.0619 | 0.3888 |
| 0.1868        | 1.1719 | 300  | 0.2227          | 0.3755 | 0.0608 | 0.3981 |
| 0.1913        | 1.5625 | 400  | 0.2274          | 0.3912 | 0.0637 | 0.3779 |
| 0.1896        | 1.9531 | 500  | 0.2263          | 0.3858 | 0.0631 | 0.3867 |
| 0.1769        | 2.3438 | 600  | 0.2176          | 0.3785 | 0.0618 | 0.3942 |
| 0.1752        | 2.7344 | 700  | 0.2162          | 0.3816 | 0.0614 | 0.3887 |
| 0.1777        | 3.125  | 800  | 0.2098          | 0.3606 | 0.0582 | 0.4260 |
| 0.1747        | 3.5156 | 900  | 0.2078          | 0.3657 | 0.0585 | 0.4111 |
| 0.1672        | 3.9062 | 1000 | 0.2075          | 0.3770 | 0.0595 | 0.3920 |
| 0.1583        | 4.2969 | 1100 | 0.2060          | 0.3631 | 0.0580 | 0.4137 |
| 0.1713        | 4.6875 | 1200 | 0.2064          | 0.3664 | 0.0587 | 0.4118 |
| 0.1563        | 5.0781 | 1300 | 0.2047          | 0.3597 | 0.0579 | 0.4226 |


### Framework versions

- Transformers 4.44.2
- Pytorch 2.4.1+cu121
- Datasets 3.0.0
- Tokenizers 0.19.1