File size: 3,641 Bytes
21144a0
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
100
101
102
103
104
105
106
107
108
109
110
111
---
license: apache-2.0
tags:
- generated_from_trainer
datasets:
- common_voice_11_0
metrics:
- wer
model-index:
- name: wav2vec2-xls-r-1b-ur
  results:
  - task:
      name: Automatic Speech Recognition
      type: automatic-speech-recognition
    dataset:
      name: common_voice_11_0
      type: common_voice_11_0
      config: ur
      split: test
      args: ur
    metrics:
    - name: Wer
      type: wer
      value: 0.48854134406937133
---

<!-- 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-xls-r-1b-ur

This model is a fine-tuned version of [facebook/wav2vec2-xls-r-1b](https://huggingface.co/facebook/wav2vec2-xls-r-1b) on the common_voice_11_0 dataset.
It achieves the following results on the evaluation set:
- Loss: inf
- Wer: 0.4885

## 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: 0.0003
- train_batch_size: 6
- eval_batch_size: 6
- seed: 42
- distributed_type: multi-GPU
- num_devices: 2
- total_train_batch_size: 12
- total_eval_batch_size: 12
- 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    |
|:-------------:|:-----:|:----:|:---------------:|:------:|
| 1.7368        | 0.48  | 300  | inf             | 0.8191 |
| 1.8995        | 0.97  | 600  | inf             | 0.7919 |
| 0.9144        | 1.45  | 900  | inf             | 0.7805 |
| 1.166         | 1.94  | 1200 | inf             | 0.7087 |
| 0.7972        | 2.42  | 1500 | inf             | 0.6901 |
| 0.8604        | 2.9   | 1800 | inf             | 0.6446 |
| 0.6569        | 3.39  | 2100 | inf             | 0.6560 |
| 0.7267        | 3.87  | 2400 | inf             | 0.6363 |
| 0.687         | 4.35  | 2700 | inf             | 0.6343 |
| 0.7143        | 4.84  | 3000 | inf             | 0.6176 |
| 0.5283        | 5.32  | 3300 | inf             | 0.6084 |
| 0.6917        | 5.81  | 3600 | inf             | 0.5942 |
| 0.5396        | 6.29  | 3900 | inf             | 0.5988 |
| 0.5523        | 6.77  | 4200 | inf             | 0.5600 |
| 0.3167        | 7.26  | 4500 | inf             | 0.5648 |
| 0.3176        | 7.74  | 4800 | inf             | 0.5424 |
| 0.3987        | 8.23  | 5100 | inf             | 0.5440 |
| 0.3327        | 8.71  | 5400 | inf             | 0.5316 |
| 0.1936        | 9.19  | 5700 | inf             | 0.5285 |
| 0.4701        | 9.68  | 6000 | inf             | 0.5207 |
| 0.3581        | 10.16 | 6300 | inf             | 0.5176 |
| 0.4038        | 10.65 | 6600 | inf             | 0.5259 |
| 0.2699        | 11.13 | 6900 | inf             | 0.5226 |
| 0.2302        | 11.61 | 7200 | inf             | 0.5181 |
| 0.3275        | 12.1  | 7500 | inf             | 0.5202 |
| 0.3024        | 12.58 | 7800 | inf             | 0.5307 |
| 0.2568        | 13.06 | 8100 | inf             | 0.5243 |
| 0.1641        | 13.55 | 8400 | inf             | 0.5073 |
| 0.2637        | 14.03 | 8700 | inf             | 0.5015 |
| 0.1778        | 14.52 | 9000 | inf             | 0.4892 |
| 0.0874        | 15.0  | 9300 | inf             | 0.4885 |


### Framework versions

- Transformers 4.27.0.dev0
- Pytorch 1.13.0
- Datasets 2.7.1
- Tokenizers 0.13.2