File size: 5,659 Bytes
7b8a22f
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
118
119
---
library_name: transformers
license: mit
base_model: openai/whisper-large-v3-turbo
tags:
- generated_from_trainer
datasets:
- fleurs
metrics:
- wer
model-index:
- name: whisper-large-v3-turbo-serbian-serbia
  results:
  - task:
      name: Automatic Speech Recognition
      type: automatic-speech-recognition
    dataset:
      name: fleurs
      type: fleurs
      config: sr_rs
      split: None
      args: sr_rs
    metrics:
    - name: Wer
      type: wer
      value: 18.343921455385505
---

<!-- 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-v3-turbo-serbian-serbia

This model is a fine-tuned version of [openai/whisper-large-v3-turbo](https://huggingface.co/openai/whisper-large-v3-turbo) on the fleurs dataset.
It achieves the following results on the evaluation set:
- Loss: 0.5289
- Model Preparation Time: 0.0069
- Wer Ortho: 34.0354
- Wer: 18.3439

## 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: 64
- eval_batch_size: 64
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 128
- optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.06
- num_epochs: 1
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch  | Step | Validation Loss | Model Preparation Time | Wer Ortho | Wer     |
|:-------------:|:------:|:----:|:---------------:|:----------------------:|:---------:|:-------:|
| 0.1827        | 0.0263 | 32   | 0.8340          | 0.0069                 | 36.3524   | 21.6142 |
| 0.1531        | 0.0527 | 64   | 0.6341          | 0.0069                 | 36.4759   | 21.7225 |
| 0.1551        | 0.0790 | 96   | 0.6893          | 0.0069                 | 36.0038   | 21.6647 |
| 0.1495        | 0.1054 | 128  | 0.5745          | 0.0069                 | 35.4518   | 21.2388 |
| 0.1388        | 0.1317 | 160  | 0.5572          | 0.0069                 | 35.5680   | 20.5241 |
| 0.1359        | 0.1581 | 192  | 0.6313          | 0.0069                 | 36.2217   | 20.6974 |
| 0.138         | 0.1844 | 224  | 0.5451          | 0.0069                 | 35.9384   | 20.4014 |
| 0.1332        | 0.2108 | 256  | 0.6109          | 0.0069                 | 35.5171   | 20.2209 |
| 0.131         | 0.2371 | 288  | 0.5748          | 0.0069                 | 35.7859   | 20.6324 |
| 0.1316        | 0.2635 | 320  | 0.5705          | 0.0069                 | 34.9942   | 20.1559 |
| 0.1296        | 0.2898 | 352  | 0.6122          | 0.0069                 | 35.3791   | 20.0332 |
| 0.1274        | 0.3162 | 384  | 0.5848          | 0.0069                 | 35.1322   | 19.8960 |
| 0.1185        | 0.3425 | 416  | 0.5458          | 0.0069                 | 35.4227   | 19.8239 |
| 0.1207        | 0.3689 | 448  | 0.5714          | 0.0069                 | 34.8053   | 19.6289 |
| 0.1193        | 0.3952 | 480  | 0.6553          | 0.0069                 | 34.7327   | 19.3546 |
| 0.119         | 0.4216 | 512  | 0.5913          | 0.0069                 | 35.0523   | 19.4629 |
| 0.122         | 0.4479 | 544  | 0.5890          | 0.0069                 | 34.8489   | 19.4846 |
| 0.1203        | 0.4743 | 576  | 0.5554          | 0.0069                 | 35.0886   | 19.5206 |
| 0.1164        | 0.5006 | 608  | 0.5637          | 0.0069                 | 34.7690   | 19.4124 |
| 0.1173        | 0.5270 | 640  | 0.5492          | 0.0069                 | 34.6601   | 19.2319 |
| 0.1175        | 0.5533 | 672  | 0.5847          | 0.0069                 | 34.4785   | 19.3474 |
| 0.1154        | 0.5797 | 704  | 0.5900          | 0.0069                 | 34.7618   | 18.8926 |
| 0.1136        | 0.6060 | 736  | 0.6104          | 0.0069                 | 34.6092   | 19.0586 |
| 0.1142        | 0.6324 | 768  | 0.5393          | 0.0069                 | 34.2533   | 18.8276 |
| 0.1156        | 0.6587 | 800  | 0.5473          | 0.0069                 | 34.5221   | 18.8926 |
| 0.1125        | 0.6851 | 832  | 0.5747          | 0.0069                 | 34.4858   | 18.8276 |
| 0.1107        | 0.7114 | 864  | 0.5470          | 0.0069                 | 34.1880   | 18.8709 |
| 0.1088        | 0.7378 | 896  | 0.5620          | 0.0069                 | 34.4785   | 18.8926 |
| 0.1111        | 0.7641 | 928  | 0.5832          | 0.0069                 | 34.4059   | 18.8493 |
| 0.1081        | 0.7904 | 960  | 0.5430          | 0.0069                 | 34.3623   | 18.8204 |
| 0.108         | 0.8168 | 992  | 0.5560          | 0.0069                 | 34.4567   | 18.7626 |
| 0.106         | 0.8431 | 1024 | 0.5344          | 0.0069                 | 34.3550   | 18.8637 |
| 0.1051        | 0.8695 | 1056 | 0.5496          | 0.0069                 | 34.4567   | 18.5894 |
| 0.1096        | 0.8958 | 1088 | 0.5341          | 0.0069                 | 34.1008   | 18.4811 |
| 0.1098        | 0.9222 | 1120 | 0.5339          | 0.0069                 | 34.2025   | 18.4306 |
| 0.1045        | 0.9485 | 1152 | 0.5289          | 0.0069                 | 34.0354   | 18.3439 |
| 0.1072        | 0.9749 | 1184 | 0.5265          | 0.0069                 | 34.0863   | 18.4089 |


### Framework versions

- Transformers 4.51.3
- Pytorch 2.5.1
- Datasets 3.6.0
- Tokenizers 0.21.1