File size: 2,621 Bytes
fcea4cc
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
---
license: apache-2.0
tags:
- generated_from_trainer
datasets:
- audiofolder
metrics:
- wer
model-index:
- name: wav2vec2-xls-r-300m-fleurs_zu-run1
  results:
  - task:
      name: Automatic Speech Recognition
      type: automatic-speech-recognition
    dataset:
      name: audiofolder
      type: audiofolder
      config: default
      split: validation
      args: default
    metrics:
    - name: Wer
      type: wer
      value: 0.600381
---

<!-- 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-300m-asr_af-run1-fleurs_zu-run1

This model is a fine-tuned version of [lucas-meyer/wav2vec2-xls-r-300m-asr_af-run1](https://huggingface.co/lucas-meyer/wav2vec2-xls-r-300m-asr_af-run1) on the audiofolder dataset.
It achieves the following results on the evaluation set:
- Loss: 0.578752
- Wer: 0.600381

## 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: 4
- eval_batch_size: 4
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- num_epochs: 30
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch | Step | Validation Loss | Wer    |
|:-------------:|:-----:|:----:|:---------------:|:------:|
| 50 | 19.607200 | 9.162900 | 1.000000 |
| 100 | 7.038300 | 4.738822 | 1.000000 |
| 150 | 4.190100 | 3.359574 | 1.000000 |
| 200 | 3.161900 | 3.032595 | 1.000000 |
| 250 | 3.004700 | 2.994741 | 1.000000 |
| 300 | 2.988300 | 2.955285 | 1.000000 |
| 350 | 2.675800 | 1.816109 | 1.000000 |
| 400 | 1.064400 | 0.866473 | 0.814220 |
| 450 | 0.601600 | 0.696754 | 0.712340 |
| 500 | 0.506900 | 0.662974 | 0.716426 |
| 550 | 0.432200 | 0.598446 | 0.667121 |
| 600 | 0.358700 | 0.618853 | 0.681013 |
| 650 | 0.333300 | 0.564290 | 0.627349 |
| 700 | 0.283100 | 0.573746 | 0.646418 |
| 750 | 0.250800 | 0.577737 | 0.639608 |
| 800 | 0.232200 | 0.557288 | 0.604467 |
| 850 | 0.191200 | 0.538959 | 0.590030 |
| 900 | 0.195600 | 0.549700 | 0.600654 |
| 950 | 0.193000 | 0.579098 | 0.611278 |
| 1000 | 0.169900 | 0.578752 | 0.600381 |


### Framework versions

- Transformers 4.28.0
- Pytorch 2.0.1+cu118
- Datasets 2.14.4
- Tokenizers 0.13.3