File size: 2,651 Bytes
195d59f
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
3d71397
195d59f
3d71397
195d59f
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
3d71397
195d59f
 
 
3d71397
195d59f
 
3d71397
 
195d59f
 
 
 
3d71397
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
195d59f
 
 
 
 
 
 
 
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
---
license: apache-2.0
base_model: facebook/wav2vec2-base
tags:
- generated_from_trainer
metrics:
- wer
model-index:
- name: Osiris_asr_model
  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. -->

# Osiris_asr_model

This model is a fine-tuned version of [facebook/wav2vec2-base](https://huggingface.co/facebook/wav2vec2-base) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 3.0600
- Wer: 1.0

## 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: 1e-05
- train_batch_size: 10
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 20
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 50
- training_steps: 1000
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch  | Step | Validation Loss | Wer    |
|:-------------:|:------:|:----:|:---------------:|:------:|
| 45.8209       | 50.0   | 50   | 21.0347         | 1.0182 |
| 10.2898       | 100.0  | 100  | 5.1552          | 1.0    |
| 5.7188        | 150.0  | 150  | 4.9140          | 1.0    |
| 5.3358        | 200.0  | 200  | 4.7650          | 1.0    |
| 5.1381        | 250.0  | 250  | 4.6797          | 1.0    |
| 4.9841        | 300.0  | 300  | 4.6168          | 1.0    |
| 4.9255        | 350.0  | 350  | 4.5741          | 1.0    |
| 4.8353        | 400.0  | 400  | 4.5321          | 1.0    |
| 4.7704        | 450.0  | 450  | 4.5100          | 1.0    |
| 4.6257        | 500.0  | 500  | 3.9382          | 1.0    |
| 3.8106        | 550.0  | 550  | 3.3939          | 1.0    |
| 3.5095        | 600.0  | 600  | 3.2887          | 1.0    |
| 3.3716        | 650.0  | 650  | 3.1967          | 1.0    |
| 3.3025        | 700.0  | 700  | 3.1539          | 1.0    |
| 3.2532        | 750.0  | 750  | 3.1477          | 1.0    |
| 3.2086        | 800.0  | 800  | 3.0984          | 1.0    |
| 3.1889        | 850.0  | 850  | 3.0857          | 1.0    |
| 3.162         | 900.0  | 900  | 3.0819          | 1.0    |
| 3.1411        | 950.0  | 950  | 3.0610          | 1.0    |
| 3.1397        | 1000.0 | 1000 | 3.0600          | 1.0    |


### Framework versions

- Transformers 4.35.2
- Pytorch 2.1.0+cu121
- Datasets 2.16.1
- Tokenizers 0.15.0