File size: 2,561 Bytes
edaa0c4
 
 
 
 
 
 
 
 
 
 
 
 
 
 
dfefcb3
edaa0c4
 
 
 
 
 
 
dfefcb3
edaa0c4
dfefcb3
edaa0c4
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
---
library_name: transformers
license: mit
base_model: openai/whisper-large-v3-turbo
tags:
- generated_from_trainer
datasets:
- common_voice_16_1
metrics:
- wer
model-index:
- name: whisper-large-v3-turbo-half
  results:
  - task:
      type: automatic-speech-recognition
      name: Automatic Speech Recognition
    dataset:
      name: common_voice_16_1
      type: common_voice_16_1
      config: en
      split: test
      args: en
    metrics:
    - type: wer
      value: 28.434990232255263
      name: Wer
---

<!-- 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-half

This model is a fine-tuned version of [openai/whisper-large-v3-turbo](https://huggingface.co/openai/whisper-large-v3-turbo) on the common_voice_16_1 dataset.
It achieves the following results on the evaluation set:
- Loss: 0.7088
- Wer: 28.4350

## 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: 0.0002
- train_batch_size: 16
- eval_batch_size: 32
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 32
- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- training_steps: 5000

### Training results

| Training Loss | Epoch | Step | Validation Loss | Wer     |
|:-------------:|:-----:|:----:|:---------------:|:-------:|
| No log        | 0     | 0    | 8.8155          | 100.0   |
| 0.9071        | 0.1   | 500  | 1.5140          | 64.0547 |
| 0.7138        | 0.2   | 1000 | 1.1375          | 49.9023 |
| 0.5078        | 0.3   | 1500 | 1.0159          | 41.3067 |
| 0.4833        | 0.4   | 2000 | 0.9379          | 34.7081 |
| 0.4164        | 0.5   | 2500 | 0.8927          | 30.9746 |
| 0.517         | 0.6   | 3000 | 0.8473          | 31.0397 |
| 0.33          | 0.7   | 3500 | 0.7714          | 27.1326 |
| 0.364         | 0.8   | 4000 | 0.7508          | 25.6132 |
| 0.3728        | 0.9   | 4500 | 0.7091          | 24.4628 |
| 0.4321        | 1.0   | 5000 | 0.7088          | 28.4350 |


### Framework versions

- Transformers 4.54.0
- Pytorch 2.8.0.dev20250319+cu128
- Datasets 3.6.0
- Tokenizers 0.21.2