File size: 1,759 Bytes
929da43
a986d1e
929da43
 
 
 
 
 
 
 
 
 
 
 
 
 
a986d1e
929da43
2477f18
 
929da43
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
2477f18
929da43
 
 
2477f18
929da43
 
 
 
 
 
 
 
 
2477f18
 
 
 
 
929da43
 
 
 
 
 
 
 
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
---
base_model: ad019el/Kabyle_xlsr-finetuned-tamasheq.en
tags:
- generated_from_trainer
metrics:
- wer
model-index:
- name: Kabyle_xlsr-finetuned-tamasheq.en
  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. -->

# Kabyle_xlsr-finetuned-tamasheq.en

This model is a fine-tuned version of [ad019el/Kabyle_xlsr-finetuned-tamasheq.en](https://huggingface.co/ad019el/Kabyle_xlsr-finetuned-tamasheq.en) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 1.5106
- Wer: 0.8739

## 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: 3e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- num_epochs: 200

### Training results

| Training Loss | Epoch | Step | Validation Loss | Wer    |
|:-------------:|:-----:|:----:|:---------------:|:------:|
| 3.637         | 3.41  | 300  | 1.9770          | 1.0348 |
| 1.1027        | 6.82  | 600  | 1.3353          | 0.8913 |
| 0.7367        | 10.23 | 900  | 1.4064          | 0.8855 |
| 0.6037        | 13.64 | 1200 | 1.3830          | 0.8768 |
| 0.529         | 17.05 | 1500 | 1.5106          | 0.8739 |


### Framework versions

- Transformers 4.32.1
- Pytorch 2.0.1+cu118
- Datasets 2.14.4
- Tokenizers 0.13.3