File size: 1,629 Bytes
b3e5294
71e42a4
32871f1
71e42a4
 
 
 
 
32871f1
71e42a4
b3e5294
 
71e42a4
 
b3e5294
32871f1
b3e5294
32871f1
71e42a4
32871f1
 
b3e5294
71e42a4
b3e5294
71e42a4
b3e5294
71e42a4
b3e5294
71e42a4
b3e5294
71e42a4
b3e5294
71e42a4
b3e5294
71e42a4
b3e5294
71e42a4
b3e5294
71e42a4
 
 
 
 
 
 
 
32871f1
71e42a4
b3e5294
71e42a4
b3e5294
71e42a4
 
32871f1
 
 
 
b3e5294
 
71e42a4
b3e5294
71e42a4
 
 
 
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
---
license: apache-2.0
base_model: facebook/wav2vec2-base
tags:
- generated_from_trainer
metrics:
- wer
model-index:
- name: 240615-wav2vec2-ASR-Chinese
  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. -->

# 240615-wav2vec2-ASR-Chinese

This model is a fine-tuned version of [facebook/wav2vec2-base](https://huggingface.co/facebook/wav2vec2-base) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 1.2831
- Wer: 0.2399

## 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: 0.0001
- train_batch_size: 5
- 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
- num_epochs: 30
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch | Step | Validation Loss | Wer    |
|:-------------:|:-----:|:----:|:---------------:|:------:|
| No log        | 6.25  | 100  | 1.2831          | 0.2399 |
| No log        | 12.5  | 200  | 1.2996          | 0.2657 |
| No log        | 18.75 | 300  | 1.3705          | 0.3092 |
| No log        | 25.0  | 400  | 1.3119          | 0.3269 |


### Framework versions

- Transformers 4.41.2
- Pytorch 2.3.1+cu121
- Datasets 2.19.2
- Tokenizers 0.19.1