---
library_name: transformers
language:
- ar
license: apache-2.0
base_model: openai/whisper-tiny
tags:
- generated_from_trainer
metrics:
- wer
model-index:
- name: Whisper tiny AR - BH
  results: []
---

<!-- 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 AR - BH

This model is a fine-tuned version of [openai/whisper-tiny](https://huggingface.co/openai/whisper-tiny) on the quran-ayat-speech-to-text dataset.
It achieves the following results on the evaluation set:
- Loss: 0.0180
- Wer: 0.1110
- Cer: 0.0427

## 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: 0.0001
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- gradient_accumulation_steps: 4
- 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_steps: 500
- num_epochs: 20
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch   | Step | Validation Loss | Wer    | Cer    |
|:-------------:|:-------:|:----:|:---------------:|:------:|:------:|
| 0.0141        | 1.0     | 157  | 0.0135          | 0.1618 | 0.0599 |
| 0.0129        | 2.0     | 314  | 0.0135          | 0.1479 | 0.0451 |
| 0.0079        | 3.0     | 471  | 0.0161          | 0.1741 | 0.0609 |
| 0.0061        | 4.0     | 628  | 0.0159          | 0.1707 | 0.0537 |
| 0.0052        | 5.0     | 785  | 0.0162          | 0.1551 | 0.0519 |
| 0.0038        | 6.0     | 942  | 0.0166          | 0.1540 | 0.0520 |
| 0.0024        | 7.0     | 1099 | 0.0173          | 0.1332 | 0.0429 |
| 0.0019        | 8.0     | 1256 | 0.0171          | 0.1419 | 0.0446 |
| 0.0013        | 9.0     | 1413 | 0.0173          | 0.1455 | 0.0514 |
| 0.0011        | 10.0    | 1570 | 0.0170          | 0.1374 | 0.0453 |
| 0.0006        | 11.0    | 1727 | 0.0166          | 0.1292 | 0.0421 |
| 0.0004        | 12.0    | 1884 | 0.0170          | 0.1314 | 0.0415 |
| 0.0003        | 13.0    | 2041 | 0.0177          | 0.1274 | 0.0407 |
| 0.0003        | 14.0    | 2198 | 0.0171          | 0.1207 | 0.0381 |
| 0.0002        | 15.0    | 2355 | 0.0173          | 0.1240 | 0.0417 |
| 0.0001        | 16.0    | 2512 | 0.0175          | 0.1137 | 0.0362 |
| 0.0           | 17.0    | 2669 | 0.0171          | 0.1106 | 0.0378 |
| 0.0           | 18.0    | 2826 | 0.0172          | 0.1015 | 0.0341 |
| 0.0           | 19.0    | 2983 | 0.0183          | 0.1225 | 0.0464 |
| 0.0           | 19.8768 | 3120 | 0.0176          | 0.1821 | 0.0720 |


### Framework versions

- Transformers 4.47.0
- Pytorch 2.5.1+cu121
- Datasets 3.2.0
- Tokenizers 0.21.0