File size: 2,263 Bytes
e45d612
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
934deb0
 
e45d612
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
934deb0
e45d612
 
 
 
 
934deb0
 
e45d612
 
 
 
 
 
934deb0
 
 
 
 
 
 
 
 
 
 
 
 
e45d612
 
 
 
 
 
 
 
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
---
license: apache-2.0
tags:
- automatic-speech-recognition
- multilingual_librispeech
- generated_from_trainer
datasets:
- multilingual_librispeech
model-index:
- name: wav2vec2-300m-mls-german-ft
  results: []
---

<!-- 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. -->

# wav2vec2-300m-mls-german-ft

This model is a fine-tuned version of [facebook/wav2vec2-xls-r-300m](https://huggingface.co/facebook/wav2vec2-xls-r-300m) on the MULTILINGUAL_LIBRISPEECH - GERMAN dataset.
It achieves the following results on the evaluation set:
- Loss: 0.2157
- Wer: 0.1562

## 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: 0.0001
- train_batch_size: 32
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 1000
- num_epochs: 100.0
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch | Step | Validation Loss | Wer    |
|:-------------:|:-----:|:----:|:---------------:|:------:|
| 3.0132        | 7.25  | 500  | 2.9393          | 1.0    |
| 2.9241        | 14.49 | 1000 | 2.8734          | 1.0    |
| 1.0766        | 21.74 | 1500 | 0.2773          | 0.2488 |
| 0.8416        | 28.99 | 2000 | 0.2224          | 0.1990 |
| 0.8048        | 36.23 | 2500 | 0.2063          | 0.1792 |
| 0.7664        | 43.48 | 3000 | 0.2088          | 0.1748 |
| 0.6571        | 50.72 | 3500 | 0.2042          | 0.1668 |
| 0.7014        | 57.97 | 4000 | 0.2136          | 0.1649 |
| 0.6171        | 65.22 | 4500 | 0.2139          | 0.1641 |
| 0.6609        | 72.46 | 5000 | 0.2144          | 0.1621 |
| 0.6318        | 79.71 | 5500 | 0.2129          | 0.1600 |
| 0.6222        | 86.96 | 6000 | 0.2124          | 0.1582 |
| 0.588         | 94.2  | 6500 | 0.2143          | 0.1560 |


### Framework versions

- Transformers 4.13.0.dev0
- Pytorch 1.10.0
- Datasets 1.15.2.dev0
- Tokenizers 0.10.3