File size: 2,648 Bytes
01cc924
 
1509327
 
01cc924
 
1509327
c260866
 
1509327
01cc924
c260866
01cc924
1509327
 
 
 
 
 
c260866
 
1509327
 
 
 
 
01cc924
 
 
 
 
75d2fba
01cc924
 
 
d9c7f3f
 
01cc924
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
---
license: apache-2.0
language:
- es
tags:
- generated_from_trainer
- robust-speech-event
- common_voice_8_0
- mozilla-foundation/common_voice_8_0

datasets:
- mozilla-foundation/common_voice_8_0
model-index:
- name: wave2vec-xls-r-300m-es
  results:
  - task: 
      name: Speech Recognition
      type: automatic-speech-recognition
    dataset:
      name: mozilla-foundation/common_voice_8_0 es
      type: mozilla-foundation/common_voice_8_0
      args: es
    metrics:
       - name: Test WER
         type: wer
         value: 14.38
---

<!-- 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. -->

# xls-r-300m-es with n-gram(5)

This model is a fine-tuned version of [facebook/wav2vec2-xls-r-300m](https://huggingface.co/facebook/wav2vec2-xls-r-300m) on the common_voice dataset.
It achieves the following results on the evaluation set:
- Loss (Test): 0.1900
- Wer (Test): 0.1438

## 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: 4
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch | Step | Validation Loss | Wer    |
|:-------------:|:-----:|:----:|:---------------:|:------:|
| 3.6747        | 0.3   | 400  | 0.6535          | 0.5926 |
| 0.4439        | 0.6   | 800  | 0.3753          | 0.3193 |
| 0.3291        | 0.9   | 1200 | 0.3267          | 0.2721 |
| 0.2644        | 1.2   | 1600 | 0.2816          | 0.2311 |
| 0.24          | 1.5   | 2000 | 0.2647          | 0.2179 |
| 0.2265        | 1.79  | 2400 | 0.2406          | 0.2048 |
| 0.1994        | 2.09  | 2800 | 0.2357          | 0.1869 |
| 0.1613        | 2.39  | 3200 | 0.2242          | 0.1821 |
| 0.1546        | 2.69  | 3600 | 0.2123          | 0.1707 |
| 0.1441        | 2.99  | 4000 | 0.2067          | 0.1619 |
| 0.1138        | 3.29  | 4400 | 0.2044          | 0.1519 |
| 0.1072        | 3.59  | 4800 | 0.1917          | 0.1457 |
| 0.0992        | 3.89  | 5200 | 0.1900          | 0.1438 |


### Framework versions

- Transformers 4.16.0.dev0
- Pytorch 1.10.1+cu102
- Datasets 1.17.1.dev0
- Tokenizers 0.11.0