---
library_name: transformers
language:
- ta
base_model: openai/whisper-tiny
tags:
- generated_from_trainer
datasets:
- ai4bharat/Shrutilipi
- deepdml/iisc-mile-tamil-asr
- fixie-ai/common_voice_17_0
- ai4bharat/Kathbath
- google/fleurs
- deepdml/microsoft-speech-corpus-indian
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: ai4bharat/Shrutilipi
    metrics:
    - name: Wer
      type: wer
      value: 51.4523837126921
---
<!-- 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: 0.2599
- Wer: 51.4524
- Cer: 11.7485

## 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: 32
- eval_batch_size: 64
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 64
- 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: 0.04
- training_steps: 8000

### Training results

| Training Loss | Epoch | Step | Validation Loss | Wer     | Cer     |
|:-------------:|:-----:|:----:|:---------------:|:-------:|:-------:|
| 0.5468        | 0.125 | 1000 | 0.3404          | 63.1256 | 15.8467 |
| 0.3868        | 0.25  | 2000 | 0.3010          | 57.6548 | 13.5049 |
| 0.3373        | 0.375 | 3000 | 0.2899          | 55.5435 | 13.1339 |
| 0.2859        | 0.5   | 4000 | 0.2790          | 53.8239 | 12.5869 |
| 0.2951        | 0.625 | 5000 | 0.2662          | 52.5924 | 12.0777 |
| 0.2761        | 0.75  | 6000 | 0.2633          | 51.9754 | 11.9105 |
| 0.2687        | 0.875 | 7000 | 0.2598          | 51.5696 | 11.7500 |
| 0.2731        | 1.0   | 8000 | 0.2599          | 51.4524 | 11.7485 |


### Framework versions

- Transformers 4.48.0.dev0
- Pytorch 2.5.1+cu121
- Datasets 3.6.0
- Tokenizers 0.21.0

## Citation

Please cite the model using the following BibTeX entry:

```bibtex
@misc{deepdml/whisper-tiny-ta-mix-norm,
      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}},
      year={2026}
    }
```