File size: 2,793 Bytes
4eb38ef
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
---
language:
- en
license: apache-2.0
base_model: openai/whisper-large-v3
tags:
- generated_from_trainer
metrics:
- wer
model-index:
- name: ./whisper-large-cit-do0.25-wd0-lr1e-06
  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. -->

# ./whisper-large-cit-do0.25-wd0-lr1e-06

This model is a fine-tuned version of [openai/whisper-large-v3](https://huggingface.co/openai/whisper-large-v3) on the SF 200 dataset.
It achieves the following results on the evaluation set:
- Loss: 0.6753
- Wer: 34.3249

## 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-06
- train_batch_size: 4
- eval_batch_size: 8
- seed: 42
- distributed_type: multi-GPU
- gradient_accumulation_steps: 4
- total_train_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 100
- training_steps: 200
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch   | Step | Validation Loss | Wer     |
|:-------------:|:-------:|:----:|:---------------:|:-------:|
| 1.1266        | 0.8889  | 10   | 1.1143          | 48.9703 |
| 1.0858        | 1.7778  | 20   | 1.0078          | 39.5881 |
| 0.9333        | 2.6667  | 30   | 0.8696          | 39.1304 |
| 0.7546        | 3.5556  | 40   | 0.7925          | 34.0961 |
| 0.7029        | 4.4444  | 50   | 0.7202          | 34.7826 |
| 0.601         | 5.3333  | 60   | 0.6558          | 32.9519 |
| 0.5097        | 6.2222  | 70   | 0.6177          | 31.5789 |
| 0.415         | 7.1111  | 80   | 0.5913          | 33.4096 |
| 0.3794        | 8.0     | 90   | 0.5728          | 32.9519 |
| 0.3026        | 8.8889  | 100  | 0.5693          | 33.4096 |
| 0.2687        | 9.7778  | 110  | 0.5732          | 33.1808 |
| 0.2175        | 10.6667 | 120  | 0.5825          | 31.8078 |
| 0.1864        | 11.5556 | 130  | 0.5942          | 33.1808 |
| 0.155         | 12.4444 | 140  | 0.6060          | 31.1213 |
| 0.1266        | 13.3333 | 150  | 0.6182          | 33.1808 |
| 0.1212        | 14.2222 | 160  | 0.6328          | 33.8673 |
| 0.1043        | 15.1111 | 170  | 0.6499          | 33.8673 |
| 0.0894        | 16.0    | 180  | 0.6616          | 32.7231 |
| 0.084         | 16.8889 | 190  | 0.6724          | 33.8673 |
| 0.0787        | 17.7778 | 200  | 0.6753          | 34.3249 |


### Framework versions

- Transformers 4.41.1
- Pytorch 1.13.1+cu117
- Datasets 2.19.1
- Tokenizers 0.19.1