File size: 2,888 Bytes
2175f23
49f165f
 
2175f23
 
49f165f
2175f23
 
49f165f
2175f23
 
 
49f165f
2175f23
 
 
 
 
49f165f
 
2175f23
 
 
 
 
 
49f165f
2175f23
 
 
 
 
49f165f
2175f23
49f165f
2175f23
49f165f
 
 
2175f23
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
85
86
87
88
89
90
91
92
93
94
95
---
language:
- zh
license: apache-2.0
tags:
- whisper-event
- generated_from_trainer
datasets:
- mozilla-foundation/common_voice_11_0
metrics:
- wer
model-index:
- name: Whisper Large Chinese (Mandarin)
  results:
  - task:
      name: Automatic Speech Recognition
      type: automatic-speech-recognition
    dataset:
      name: mozilla-foundation/common_voice_11_0 zh-CN
      type: mozilla-foundation/common_voice_11_0
      config: zh-CN
      split: validation[:1000]
      args: zh-CN
    metrics:
    - name: Wer
      type: wer
      value: 51.67420814479639
---

<!-- 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 Large Chinese (Mandarin)

This model is a fine-tuned version of [openai/whisper-large-v2](https://huggingface.co/openai/whisper-large-v2) on the mozilla-foundation/common_voice_11_0 zh-CN dataset.
It achieves the following results on the evaluation set:
- Loss: 0.2435
- Wer: 51.6742
- Cer: 8.5279

## 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: 5e-06
- 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: 2000
- training_steps: 20000
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch | Step  | Validation Loss | Wer     | Cer     |
|:-------------:|:-----:|:-----:|:---------------:|:-------:|:-------:|
| 0.3314        | 0.83  | 1000  | 0.2110          | 65.7014 | 10.8047 |
| 0.2747        | 1.66  | 2000  | 0.2005          | 58.1900 | 9.4191  |
| 0.1989        | 2.49  | 3000  | 0.1983          | 56.1991 | 9.0939  |
| 0.1142        | 3.31  | 4000  | 0.2076          | 55.0226 | 9.1589  |
| 0.0747        | 4.14  | 5000  | 0.2131          | 56.3801 | 9.0483  |
| 0.0709        | 4.97  | 6000  | 0.2165          | 54.6606 | 8.9768  |
| 0.0432        | 5.8   | 7000  | 0.2222          | 54.0271 | 8.9508  |
| 0.0261        | 6.63  | 8000  | 0.2299          | 54.4796 | 9.0353  |
| 0.0152        | 7.46  | 9000  | 0.2290          | 52.7602 | 8.8076  |
| 0.0054        | 8.28  | 10000 | 0.2435          | 51.6742 | 8.5279  |
| 0.0028        | 9.11  | 11000 | 0.2421          | 53.0317 | 8.9833  |
| 0.0045        | 9.94  | 12000 | 0.2462          | 52.9412 | 8.7751  |
| 0.0016        | 10.77 | 13000 | 0.2501          | 52.3077 | 8.9573  |


### Framework versions

- Transformers 4.26.0.dev0
- Pytorch 1.13.1+cu117
- Datasets 2.7.1.dev0
- Tokenizers 0.13.2