File size: 4,595 Bytes
f2f8ea0
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
---
license: apache-2.0
base_model: openai/whisper-tiny
tags:
- generated_from_trainer
datasets:
- PolyAI/minds14
metrics:
- wer
model-index:
- name: whisper-tiny-en-US
  results:
  - task:
      name: Automatic Speech Recognition
      type: automatic-speech-recognition
    dataset:
      name: PolyAI/minds14
      type: PolyAI/minds14
      config: en-AU
      split: train
      args: en-AU
    metrics:
    - name: Wer
      type: wer
      value: 0.1655499720826354
---

<!-- 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-tiny-en-US

This model is a fine-tuned version of [openai/whisper-tiny](https://huggingface.co/openai/whisper-tiny) on the PolyAI/minds14 dataset.
It achieves the following results on the evaluation set:
- Loss: 0.4245
- Wer Ortho: 0.1714
- Wer: 0.1655

## 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: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: constant_with_warmup
- lr_scheduler_warmup_steps: 5
- training_steps: 400

### Training results

| Training Loss | Epoch | Step | Validation Loss | Wer Ortho | Wer    |
|:-------------:|:-----:|:----:|:---------------:|:---------:|:------:|
| No log        | 0.36  | 10   | 3.1022          | 0.3282    | 0.1960 |
| No log        | 0.71  | 20   | 1.6867          | 0.2399    | 0.1865 |
| 2.9245        | 1.07  | 30   | 0.6685          | 0.2332    | 0.1982 |
| 2.9245        | 1.43  | 40   | 0.4912          | 0.2017    | 0.1848 |
| 0.6297        | 1.79  | 50   | 0.4243          | 0.1865    | 0.1753 |
| 0.6297        | 2.14  | 60   | 0.3895          | 0.1801    | 0.1689 |
| 0.6297        | 2.5   | 70   | 0.3678          | 0.1769    | 0.1669 |
| 0.3045        | 2.86  | 80   | 0.3570          | 0.1746    | 0.1689 |
| 0.3045        | 3.21  | 90   | 0.3496          | 0.1720    | 0.1647 |
| 0.1949        | 3.57  | 100  | 0.3451          | 0.1746    | 0.1661 |
| 0.1949        | 3.93  | 110  | 0.3407          | 0.1804    | 0.1700 |
| 0.1949        | 4.29  | 120  | 0.3439          | 0.1778    | 0.1695 |
| 0.1099        | 4.64  | 130  | 0.3501          | 0.1743    | 0.1689 |
| 0.1099        | 5.0   | 140  | 0.3488          | 0.1737    | 0.1667 |
| 0.0583        | 5.36  | 150  | 0.3554          | 0.1778    | 0.1697 |
| 0.0583        | 5.71  | 160  | 0.3595          | 0.1708    | 0.1628 |
| 0.0583        | 6.07  | 170  | 0.3514          | 0.1746    | 0.1661 |
| 0.032         | 6.43  | 180  | 0.3672          | 0.1755    | 0.1683 |
| 0.032         | 6.79  | 190  | 0.3676          | 0.1676    | 0.1602 |
| 0.0146        | 7.14  | 200  | 0.3791          | 0.1658    | 0.1600 |
| 0.0146        | 7.5   | 210  | 0.3825          | 0.1676    | 0.1625 |
| 0.0146        | 7.86  | 220  | 0.3799          | 0.1702    | 0.1650 |
| 0.0084        | 8.21  | 230  | 0.3827          | 0.1702    | 0.1655 |
| 0.0084        | 8.57  | 240  | 0.3869          | 0.1778    | 0.1714 |
| 0.0043        | 8.93  | 250  | 0.3951          | 0.1740    | 0.1686 |
| 0.0043        | 9.29  | 260  | 0.3958          | 0.1720    | 0.1672 |
| 0.0043        | 9.64  | 270  | 0.3968          | 0.1758    | 0.1706 |
| 0.003         | 10.0  | 280  | 0.3978          | 0.1725    | 0.1672 |
| 0.003         | 10.36 | 290  | 0.4012          | 0.1734    | 0.1681 |
| 0.0023        | 10.71 | 300  | 0.4068          | 0.1728    | 0.1678 |
| 0.0023        | 11.07 | 310  | 0.4097          | 0.1752    | 0.1697 |
| 0.0023        | 11.43 | 320  | 0.4113          | 0.1746    | 0.1692 |
| 0.0018        | 11.79 | 330  | 0.4120          | 0.1737    | 0.1681 |
| 0.0018        | 12.14 | 340  | 0.4141          | 0.1740    | 0.1683 |
| 0.0016        | 12.5  | 350  | 0.4172          | 0.1731    | 0.1678 |
| 0.0016        | 12.86 | 360  | 0.4193          | 0.1740    | 0.1681 |
| 0.0016        | 13.21 | 370  | 0.4197          | 0.1731    | 0.1672 |
| 0.0014        | 13.57 | 380  | 0.4215          | 0.1731    | 0.1672 |
| 0.0014        | 13.93 | 390  | 0.4228          | 0.1720    | 0.1664 |
| 0.0012        | 14.29 | 400  | 0.4245          | 0.1714    | 0.1655 |


### Framework versions

- Transformers 4.31.0.dev0
- Pytorch 2.0.1+cu118
- Datasets 2.13.1
- Tokenizers 0.13.3