File size: 2,046 Bytes
9cd2101
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
20da83b
9cd2101
 
 
 
 
 
 
 
 
20da83b
 
9cd2101
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
20da83b
9cd2101
 
20da83b
9cd2101
 
 
 
20da83b
9cd2101
 
 
 
20da83b
 
 
 
 
 
9cd2101
 
 
 
 
 
 
 
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
---
language:
- hr
license: apache-2.0
base_model: openai/whisper-small
tags:
- generated_from_trainer
datasets:
- parlaSmall
metrics:
- wer
model-index:
- name: Whisper Small Croatian
  results:
  - task:
      name: Automatic Speech Recognition
      type: automatic-speech-recognition
    dataset:
      name: parlaSmall_subset
      type: parlaSmall
      config: default
      split: None
      args: 'config: hr, split: test'
    metrics:
    - name: Wer
      type: wer
      value: 25.440806045340054
---

<!-- 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 Small Croatian

This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the parlaSmall_subset dataset.
It achieves the following results on the evaluation set:
- Loss: 0.5739
- Wer: 25.4408

## 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: 2
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 8
- 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: 500
- training_steps: 4000
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch  | Step | Validation Loss | Wer     |
|:-------------:|:------:|:----:|:---------------:|:-------:|
| 0.0003        | 32.52  | 1000 | 0.5073          | 25.0630 |
| 0.0001        | 65.04  | 2000 | 0.5470          | 25.5668 |
| 0.0001        | 97.56  | 3000 | 0.5668          | 25.0630 |
| 0.0           | 130.08 | 4000 | 0.5739          | 25.4408 |


### Framework versions

- Transformers 4.39.3
- Pytorch 2.0.0
- Datasets 2.19.1
- Tokenizers 0.15.1