metadata

library_name: transformers
language:
  - ar
license: apache-2.0
base_model: openai/whisper-small
tags:
  - automatic-speech-recognition
  - arabic
  - generated_from_trainer
datasets:
  - itskavya/gp
metrics:
  - wer
model-index:
  - name: Whisper Small Informal Arabic
    results:
      - task:
          name: Automatic Speech Recognition
          type: automatic-speech-recognition
        dataset:
          name: Informal Arabic
          type: itskavya/gp
          config: default
          split: train
          args: default
        metrics:
          - name: Wer
            type: wer
            value: 31.308703819661865

Whisper Small Informal Arabic

This model is a fine-tuned version of openai/whisper-small on the Informal Arabic dataset. It achieves the following results on the evaluation set:

Loss: 0.5477
Wer: 31.3087

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: 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: 100
training_steps: 2000
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.0003	71.4286	1000	0.5191	30.3694
0.0002	142.8571	2000	0.5477	31.3087

Framework versions

Transformers 4.49.0
Pytorch 2.5.1+cu118
Datasets 3.3.2
Tokenizers 0.21.0