---
library_name: transformers
language:
- ar
license: apache-2.0
base_model: openai/whisper-small
tags:
- generated_from_trainer
datasets:
- ZC/fintunewhisperarab
metrics:
- wer
model-index:
- name: Whisper Small arab
  results:
  - task:
      name: Automatic Speech Recognition
      type: automatic-speech-recognition
    dataset:
      name: arabcorpus
      type: ZC/fintunewhisperarab
      args: 'config: ar, split: test'
    metrics:
    - name: Wer
      type: wer
      value: 34.60026212319791
---

<!-- 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 arab

This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the arabcorpus dataset.
It achieves the following results on the evaluation set:
- Loss: 0.2144
- Wer: 34.6003

## 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: 8
- seed: 42
- 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
- training_steps: 1000
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch  | Step | Validation Loss | Wer     |
|:-------------:|:------:|:----:|:---------------:|:-------:|
| 0.021         | 4.1494 | 1000 | 0.2144          | 34.6003 |


### Framework versions

- Transformers 4.46.3
- Pytorch 2.5.1+cu121
- Datasets 3.1.0
- Tokenizers 0.20.3