File size: 1,698 Bytes
7a94994
 
 
 
3f72131
7a94994
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
---
license: apache-2.0
tags:
- generated_from_trainer
base_model: ad019el/tamasheq-98-1
model-index:
- name: tamasheq-98-2
  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. -->

# tamasheq-98-2

This model is a fine-tuned version of [ad019el/tamasheq-98-1](https://huggingface.co/ad019el/tamasheq-98-1) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 1.5349
- Wer: 0.9667

## 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: 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: 200
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch | Step | Validation Loss | Wer    |
|:-------------:|:-----:|:----:|:---------------:|:------:|
| 7.6372        | 9.62  | 500  | 2.3244          | 1.0519 |
| 0.5009        | 19.23 | 1000 | 1.4651          | 0.9444 |
| 0.2431        | 28.85 | 1500 | 1.5561          | 0.9741 |
| 0.2013        | 38.46 | 2000 | 1.6960          | 1.0407 |
| 0.1857        | 48.08 | 2500 | 1.5349          | 0.9667 |


### Framework versions

- Transformers 4.20.1
- Pytorch 1.11.0
- Datasets 2.1.0
- Tokenizers 0.12.1