File size: 4,371 Bytes
fdafed2
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
112
113
114
115
---
tags:
- generated_from_trainer
datasets:
- evanarlian/common_voice_11_0_id_filtered
metrics:
- wer
model-index:
- name: wav2vec2-xls-r-164m-id
  results:
  - task:
      name: Automatic Speech Recognition
      type: automatic-speech-recognition
    dataset:
      name: evanarlian/common_voice_11_0_id_filtered
      type: evanarlian/common_voice_11_0_id_filtered
    metrics:
    - name: Wer
      type: wer
      value: 0.3199428097039019
---

<!-- 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-164m-id

This model is a fine-tuned version of [evanarlian/distil-wav2vec2-xls-r-164m-id](https://huggingface.co/evanarlian/distil-wav2vec2-xls-r-164m-id) on the evanarlian/common_voice_11_0_id_filtered dataset.
It achieves the following results on the evaluation set:
- Loss: 0.3215
- Wer: 0.3199

## 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: 0.0002
- train_batch_size: 24
- eval_batch_size: 24
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.3
- num_epochs: 40.0
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch | Step  | Validation Loss | Wer    |
|:-------------:|:-----:|:-----:|:---------------:|:------:|
| 3.5445        | 0.92  | 1000  | 3.0106          | 1.0000 |
| 2.5067        | 1.84  | 2000  | 1.6134          | 0.9905 |
| 1.0279        | 2.75  | 3000  | 0.7667          | 0.8217 |
| 0.7823        | 3.67  | 4000  | 0.6141          | 0.7224 |
| 0.6504        | 4.59  | 5000  | 0.5228          | 0.6503 |
| 0.5687        | 5.51  | 6000  | 0.4666          | 0.5963 |
| 0.5026        | 6.43  | 7000  | 0.4288          | 0.5612 |
| 0.4584        | 7.35  | 8000  | 0.4048          | 0.5267 |
| 0.4193        | 8.26  | 9000  | 0.4057          | 0.5218 |
| 0.3931        | 9.18  | 10000 | 0.3820          | 0.4813 |
| 0.3651        | 10.1  | 11000 | 0.3686          | 0.4709 |
| 0.3526        | 11.02 | 12000 | 0.3665          | 0.4655 |
| 0.3333        | 11.94 | 13000 | 0.3440          | 0.4485 |
| 0.3095        | 12.86 | 14000 | 0.3314          | 0.4331 |
| 0.2802        | 13.77 | 15000 | 0.3360          | 0.4157 |
| 0.2724        | 14.69 | 16000 | 0.3331          | 0.4107 |
| 0.2488        | 15.61 | 17000 | 0.3255          | 0.4037 |
| 0.231         | 16.53 | 18000 | 0.3089          | 0.3950 |
| 0.2146        | 17.45 | 19000 | 0.3398          | 0.3990 |
| 0.2103        | 18.37 | 20000 | 0.3080          | 0.3805 |
| 0.2035        | 19.28 | 21000 | 0.3158          | 0.3828 |
| 0.1933        | 20.2  | 22000 | 0.3118          | 0.3728 |
| 0.1839        | 21.12 | 23000 | 0.3076          | 0.3690 |
| 0.1791        | 22.04 | 24000 | 0.3041          | 0.3658 |
| 0.1696        | 22.96 | 25000 | 0.3092          | 0.3603 |
| 0.1608        | 23.88 | 26000 | 0.2936          | 0.3555 |
| 0.1568        | 24.79 | 27000 | 0.2936          | 0.3560 |
| 0.1456        | 25.71 | 28000 | 0.3257          | 0.3543 |
| 0.1399        | 26.63 | 29000 | 0.3100          | 0.3424 |
| 0.1345        | 27.55 | 30000 | 0.3172          | 0.3472 |
| 0.1264        | 28.47 | 31000 | 0.3276          | 0.3412 |
| 0.1289        | 29.38 | 32000 | 0.3104          | 0.3401 |
| 0.1246        | 30.3  | 33000 | 0.3204          | 0.3352 |
| 0.1156        | 31.22 | 34000 | 0.3013          | 0.3353 |
| 0.1143        | 32.14 | 35000 | 0.3102          | 0.3322 |
| 0.1152        | 33.06 | 36000 | 0.3240          | 0.3323 |
| 0.1093        | 33.98 | 37000 | 0.3105          | 0.3295 |
| 0.101         | 34.89 | 38000 | 0.3112          | 0.3263 |
| 0.1017        | 35.81 | 39000 | 0.3263          | 0.3239 |
| 0.0915        | 36.73 | 40000 | 0.3176          | 0.3226 |
| 0.0943        | 37.65 | 41000 | 0.3141          | 0.3210 |
| 0.0898        | 38.57 | 42000 | 0.3177          | 0.3183 |
| 0.0923        | 39.49 | 43000 | 0.3215          | 0.3199 |


### Framework versions

- Transformers 4.27.0.dev0
- Pytorch 1.13.1+cu117
- Datasets 2.9.1.dev0
- Tokenizers 0.13.2