File size: 2,264 Bytes
6a951b7
 
 
 
258d745
 
 
6a951b7
 
258d745
 
6a951b7
 
 
 
 
258d745
6a951b7
 
 
 
 
 
 
258d745
6a951b7
258d745
6a951b7
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
---
base_model: openai/whisper-large-v3
datasets:
- mozilla-foundation/common_voice_17_0
language:
- hu
license: apache-2.0
metrics:
- wer
tags:
- generated_from_trainer
model-index:
- name: Whisper Large V3 HU Full -  snoopyben27
  results:
  - task:
      type: automatic-speech-recognition
      name: Automatic Speech Recognition
    dataset:
      name: Common Voice 17.0
      type: mozilla-foundation/common_voice_17_0
      config: default
      split: test
      args: 'config: hu, split: test'
    metrics:
    - type: wer
      value: 8.860932585806099
      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 HU Full -  snoopyben27

This model is a fine-tuned version of [openai/whisper-large-v3](https://huggingface.co/openai/whisper-large-v3) on the Common Voice 17.0 dataset.
It achieves the following results on the evaluation set:
- Loss: 0.0911
- Wer: 8.8609

## 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: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- training_steps: 7000
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch  | Step | Validation Loss | Wer     |
|:-------------:|:------:|:----:|:---------------:|:-------:|
| 0.1301        | 0.3299 | 1000 | 0.1351          | 14.5084 |
| 0.1324        | 0.6598 | 2000 | 0.1208          | 13.2777 |
| 0.1136        | 0.9898 | 3000 | 0.1066          | 11.5548 |
| 0.0471        | 1.3197 | 4000 | 0.1030          | 10.3788 |
| 0.0337        | 1.6496 | 5000 | 0.0955          | 9.8045  |
| 0.0311        | 1.9795 | 6000 | 0.0875          | 9.2438  |
| 0.0108        | 2.3095 | 7000 | 0.0911          | 8.8609  |


### Framework versions

- Transformers 4.42.2
- Pytorch 2.3.1
- Datasets 2.20.0
- Tokenizers 0.19.1