File size: 2,681 Bytes
010408b
62d578c
 
8efa8b9
62d578c
 
 
 
 
 
 
 
 
8efa8b9
62d578c
 
 
 
 
 
 
8efa8b9
62d578c
8efa8b9
010408b
 
62d578c
 
010408b
62d578c
010408b
62d578c
 
 
 
010408b
62d578c
010408b
62d578c
010408b
62d578c
010408b
62d578c
010408b
62d578c
010408b
62d578c
010408b
62d578c
010408b
62d578c
010408b
62d578c
 
 
 
 
 
 
 
 
 
 
 
010408b
62d578c
010408b
62d578c
 
 
 
 
 
 
 
 
 
 
 
 
 
 
010408b
 
62d578c
010408b
62d578c
 
 
 
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
---
tags:
- generated_from_trainer
base_model: batoula187/wav2vec2-large-xls-r-300m-arabic-colab
datasets:
- common_voice_12_0
metrics:
- wer
model-index:
- name: wav2vec2-large-xls-r-300m-arabic-colab
  results:
  - task:
      type: automatic-speech-recognition
      name: Automatic Speech Recognition
    dataset:
      name: common_voice_12_0
      type: common_voice_12_0
      config: ar
      split: test[:10%]
      args: ar
    metrics:
    - type: wer
      value: 0.7661710754972002
      name: Wer
---

<!-- 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-xls-r-300m-arabic-colab

This model is a fine-tuned version of [batoula187/wav2vec2-large-xls-r-300m-arabic-colab](https://huggingface.co/batoula187/wav2vec2-large-xls-r-300m-arabic-colab) on the common_voice_12_0 dataset.
It achieves the following results on the evaluation set:
- Loss: 2.0728
- Wer: 0.7662

## 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: 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: 30
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch   | Step | Validation Loss | Wer    |
|:-------------:|:-------:|:----:|:---------------:|:------:|
| 0.1061        | 2.2599  | 200  | 1.8297          | 0.8034 |
| 0.1496        | 4.5198  | 400  | 1.6173          | 0.7955 |
| 0.2105        | 6.7797  | 600  | 1.6220          | 0.8040 |
| 0.1798        | 9.0395  | 800  | 2.2087          | 0.8405 |
| 0.1389        | 11.2994 | 1000 | 1.7900          | 0.7868 |
| 0.1143        | 13.5593 | 1200 | 1.7566          | 0.7886 |
| 0.103         | 15.8192 | 1400 | 1.8148          | 0.7689 |
| 0.0904        | 18.0791 | 1600 | 1.8059          | 0.7627 |
| 0.0766        | 20.3390 | 1800 | 2.1398          | 0.7907 |
| 0.0682        | 22.5989 | 2000 | 2.0384          | 0.7779 |
| 0.0583        | 24.8588 | 2200 | 2.0727          | 0.7658 |
| 0.0575        | 27.1186 | 2400 | 2.1649          | 0.7758 |
| 0.0582        | 29.3785 | 2600 | 2.0728          | 0.7662 |


### Framework versions

- Transformers 4.40.2
- Pytorch 2.2.1+cu121
- Datasets 2.19.1
- Tokenizers 0.19.1