README.md

---
language:
- ar
license: apache-2.0
tags:
- hf-asr-leaderboard
- generated_from_trainer
datasets:
- mozilla-foundation/common_voice_11_0
metrics:
- wer
model-index:
- name: Whisper Small Ar-Martha
  results:
  - task:
      name: Automatic Speech Recognition
      type: automatic-speech-recognition
    dataset:
      name: Common Voice 11.0
      type: mozilla-foundation/common_voice_11_0
      args: 'config: ar'
    metrics:
    - name: Wer
      type: wer
      value: 70.20710621318639
---

<!-- 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 Ar- Martha:
  
  This model is a fine-tuned version of openai/whisper-small on the Common Voice 11.0 dataset. It achieves the following results on the evaluation set:

Loss: 0.5854

Wer: 70.2071

## 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: 16
 
eval_batch_size: 8

seed: 42

optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08

lr_scheduler_type: linear

lr_scheduler_warmup_steps: 500

training_steps: 500

mixed_precision_training: Native AMP

# Training results

| Training Loss | Epoch | Step | Validation Loss | Wer     |
|:-------------:|:-----:|:----:|:---------------:|:-------:|
| 0.9692        | 0.14  | 125  | 1.3372          | 173.0952|
| 0.5716        | 0.29  | 250  | 0.9058          | 148.6795|
| 0.3297        | 0.43  | 375  | 0.5825          | 63.6709 |
| 0.3083	    | 0.57  | 500  | 0.5854          | 70.2071 |



## Framework versions

Transformers 4.26.0.dev0

Pytorch 1.13.0+cu116

Datasets 2.7.1

Tokenizers 0.13.2