Hugging Face's logo

deepdml
/
whisper-tiny-ta-mix-norm-optim

TensorBoard
Safetensors
Tamil
whisper
Generated from Trainer
Eval Results (legacy)
Model card Files
xet
Metrics Community
whisper-tiny-ta-mix-norm-optim
File size: 2,411 Bytes
23d14a1
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
db4c5db
 
 
 
 
 
 
 
 
 
 
 
 
 
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
 
---
language:
- ta
license: apache-2.0
base_model: openai/whisper-tiny
tags:
- generated_from_trainer
datasets:
- fixie-ai/common_voice_17_0
- google/fleurs
metrics:
- wer
model-index:
- name: Whisper Tiny ta
  results:
  - task:
      name: Automatic Speech Recognition
      type: automatic-speech-recognition
    dataset:
      name: Common Voice 17.0
      type: fixie-ai/common_voice_17_0
    metrics:
    - name: Wer
      type: wer
      value: 75.4221895892105
---
<!-- 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 Tiny ta

This model is a fine-tuned version of [openai/whisper-tiny](https://huggingface.co/openai/whisper-tiny) on the Common Voice 17.0 dataset.
It achieves the following results on the evaluation set:
- Loss: 0.4798
- Wer: 75.4222
- Cer: 21.7188

## 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
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.04
- training_steps: 5000

### Training results

| Training Loss | Epoch  | Step | Validation Loss | Wer     | Cer     |
|:-------------:|:------:|:----:|:---------------:|:-------:|:-------:|
| 0.4252        | 0.2    | 1000 | 0.5015          | 76.9190 | 22.6354 |
| 0.4034        | 0.4    | 2000 | 0.4818          | 75.6283 | 21.8341 |
| 0.4708        | 1.1734 | 3000 | 0.4798          | 75.1697 | 21.4852 |
| 0.4435        | 1.3734 | 4000 | 0.4797          | 75.3011 | 21.6158 |
| 0.457         | 2.1468 | 5000 | 0.4798          | 75.4222 | 21.7188 |


### Framework versions

- Transformers 4.42.0.dev0
- Pytorch 2.3.0+cu121
- Datasets 2.19.1
- Tokenizers 0.19.1

## Citation

Please cite the model using the following BibTeX entry:

```bibtex
@misc{deepdml/whisper-tiny-ta-mix-norm-opti,
      title={Fine-tuned Whisper tiny ASR model for speech recognition in Tamil},
      author={Jimenez, David},
      howpublished={\url{https://huggingface.co/deepdml/whisper-tiny-ta-mix-norm-opti}},
      year={2026}
    }
```