File size: 1,532 Bytes
26c3a37
e451dbf
 
 
bbf05f6
 
894e602
 
e451dbf
 
bbf05f6
 
 
26c3a37
bbf05f6
 
 
 
 
 
894e602
 
5674454
 
bbf05f6
 
 
 
 
 
 
 
 
 
 
d52a8bd
bbf05f6
 
 
 
 
 
 
 
 
 
 
 
 
 
894e602
 
 
 
5674454
894e602
 
bbf05f6
 
 
 
 
d52a8bd
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
---
language:
- wo
license: apache-2.0
tags:
- generated_from_trainer
metrics:
- wer
pipeline_tag: automatic-speech-recognition
base_model: cifope/whisper-small-wolof
model-index:
- name: whisper-small-wolof
  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. -->

# whisper-small-wolof

This model is a fine-tuned version of [cifope/whisper-small-wolof](https://huggingface.co./cifope/whisper-small-wolof) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 0.501167
- Wer: 27.1604

## Model description

More information needed

## Intended uses & limitations

More information needed

## Training and evaluation data

the model was trained on 9000 examples and tested on 1500 examples

### Training hyperparameters

The following hyperparameters were used during training:
- learning_rate: 1e-05
- train_batch_size: 16
- 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: 500
- training_steps: 1000
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch | Step | Validation Loss | Wer     |
|:-------------:|:-----:|:----:|:---------------:|:-------:|
| 0.0071        | 5.32  | 1000 | 0.501167        | 27.1604 |


### Framework versions

- Transformers 4.29.2
- Pytorch 2.0.1+cu117
- Datasets 2.12.0
- Tokenizers 0.13.3